25 Oct
2023
25 Oct
'23
4:50 a.m.
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to rebuild the
entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was alway marked
as
“experimental” and I never tried to get it running. If you need a DECnet router,
pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux versions
were changes to the kernel framework APIs. By limiting the use of these APIs I am
hoping
that the kernel module will need fewer changes to keep up with kernel changes. So far,
I have had to make one additional conditional code change at kernel 6.5 and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal” is still
available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management listener.
- There is now a subset implementation of “ncp” which is sufficient to support all of the
requests
available from the Linux network management listener. It does implement a “tell” prefix
so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node name
database.
- The new kernel module now supports node counters as defined in the Network Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi releases. The
current
procedure supports installation on systems derived from Debian or Fedora although only
a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
25 Oct
25 Oct
10:23 a.m.
Ubuntu 20.04 works fine. Thank you, that is what I have been searching for
a long time.
On Wed, Oct 25, 2023 at 4:50 AM John Forecast <john(a)forecast.name> wrote:
Sometime around mid-2022 the Linux kernel developers decided to remove the
DECnet code from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a
DECnet implementation. More
recently, there has been some discussion around removing the code from the
Long Term Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be
installed on recent releases.
Over the past year or so, I have been working on a replacement for this
repository with the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no
longer need to rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The
routing code was alway marked as
“experimental” and I never tried to get it running. If you
need a DECnet router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running
between Linux versions
were changes to the kernel framework APIs. By limiting the
use of these APIs I am hoping
that the kernel module will need fewer changes to keep up
with kernel changes. So far,
I have had to make one additional conditional code change
at kernel 6.5 and that was in
the socket layer so all networking code would need to be
changed.
In addition to a new kernel module there have been a number of changes to
the userland code:
- What used to be “fal2” is now the default file access listener.
The old “fal” is still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network
management listener.
- There is now a subset implementation of “ncp” which is
sufficient to support all of the requests
available from the Linux network management listener. It does
implement a “tell” prefix so all
of these commands may be issued to remote systems. In addition,
it implements
“ncp copy known nodes from <node name/address>” to update the
local node name database.
- The new kernel module now supports node counters as defined in
the Network Management
Specification and the ncp/nml combination are able to display
and zero them.
- The installation procedure is no longer targeted solely at
Raspberry Pi releases. The current
procedure supports installation on systems derived from Debian
or Fedora although only a
limited number of distributions have been tested (see
README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet>
and should be considered to be in Beta test at this time. See
README.DECnet in the top level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
4:47 p.m.
I successfully installed Linux DECnet (beta test) on Ubuntu 23.10 but have some problems
with that.
I executed 'sudo ./BuildAndInstall.sh' and followed instructions to enter, etc.
Also I started pydecnet (route to HECnet network) as bridge.
I tried to set host to my VAX simulator but got no route to host errors.
Also I tried to access 1.1:: but got no route host errors as well. Look that below.
dndir 1.1::
connect failed: No route to host
sethost tsvmsb
sethost V1.0.4
Connecting to tsvmsb
connect (cterm): No route to host
ct_setup_link: error 0 connecting to host
sudo ncp show known nodes
Version mismatch
sudo ncp copy known nodes from 1.1
Unable to connect to listener
lsmod | grep decnet3
decnet3 86016 5
Do you have any solutions with those problems? On my VAX simulator (TSVMSB), I was able
access HECnet without any problems.
I tried to set host to Ubuntu 23.10 but immediately got error - network is unreachable.
Tim
________________________________
From: John Forecast <john(a)forecast.name>
Sent: Tuesday, October 24, 2023 10:50 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to
rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was
alway marked as
“experimental” and I never tried to get it running. If you need a DECnet
router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux
versions
were changes to the kernel framework APIs. By limiting the use of these
APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel
changes. So far,
I have had to make one additional conditional code change at kernel 6.5
and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal” is
still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management listener.
- There is now a subset implementation of “ncp” which is sufficient to support all
of the requests
available from the Linux network management listener. It does implement a “tell”
prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node
name database.
- The new kernel module now supports node counters as defined in the Network
Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi
releases. The current
procedure supports installation on systems derived from Debian or Fedora
although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
6:17 p.m.
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
On Oct 25, 2023, at 10:47 AM, Timothy Stark
<fsword007(a)gmail.com> wrote:
I successfully installed Linux DECnet (beta test) on Ubuntu 23.10 but have some problems
with that.
I executed 'sudo ./BuildAndInstall.sh' and followed instructions to enter, etc.
Also I started pydecnet (route to HECnet network) as bridge.
I tried to set host to my VAX simulator but got no route to host errors.
Also I tried to access 1.1:: but got no route host errors as well. Look that below.
dndir 1.1::
connect failed: No route to host
I see this as well.
sethost tsvmsb
sethost V1.0.4
Connecting to tsvmsb
connect (cterm): No route to host
ct_setup_link: error 0 connecting to host
I don’t have a VMS system set up to try this but I get further trying to connect to
RSX-11m+:
forecast@emulat:~$ sethost 41.200
sethost V1.0.4
Connecting to 41.200
ct_setup_link: error sending init sequence
sudo ncp show known nodes
Version mismatch
You don’t need the “sudo” for normal operation. The “Version mismatch” is in response to
the connection completing with the wrong
optional data. Did this system have the previous DECnet for Linux installed? If so, did
you follow the section in README.DECnet
about deleting old libraries.
sudo ncp copy known nodes from 1.1
Unable to connect to listener
Same problem as dndir above.
lsmod | grep decnet3
decnet3 86016 5
That looks good. What about the output from “systemctl status decnet3”
John.
Do you have any solutions with those problems? On my
VAX simulator (TSVMSB), I was able access HECnet without any problems.
I tried to set host to Ubuntu 23.10 but immediately got error - network is unreachable.
Tim
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Tuesday, October 24, 2023 10:50 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet
<https://github.com/JohnForecast/RaspbianDECnet>> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to
rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was
alway marked as
“experimental” and I never tried to get it running. If you need a DECnet
router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux
versions
were changes to the kernel framework APIs. By limiting the use of these
APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel
changes. So far,
I have had to make one additional conditional code change at kernel 6.5
and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal”
is still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management
listener.
- There is now a subset implementation of “ncp” which is sufficient to support
all of the requests
available from the Linux network management listener. It does implement a
“tell” prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node
name database.
- The new kernel module now supports node counters as defined in the Network
Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi
releases. The current
procedure supports installation on systems derived from Debian or Fedora
although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet
<https://github.com/JohnForecast/LinuxDECnet>>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
6:23 p.m.
On Oct 25, 2023, at 12:17 PM, John Forecast
<john(a)forecast.name> wrote:
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
According to the HECnet mapper (http://akdesign.dyndns.org:8080/map/data) MAGICA has been
down since September 2021.
paul
6:34 p.m.
On Oct 25, 2023, at 12:17 PM, John Forecast
<john(a)forecast.name> wrote:
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
What does it mean for nodes to be displayed in that list? For example, 28NH isn't
connected by Ethernet (it is an endnode with a DDCMP link). So it isn't a neighbor
and that hardware address isn't meaningful.
paul
6:43 p.m.
On Oct 25, 2023, at 12:34 PM, Paul Koning
<paulkoning(a)comcast.net> wrote:
That’s actually coming from the next-hop cache. The address should actually be labeled
“Next-Hop” and contain 41.250’s MAC address. It’s a leftover from the old implementation.
John.
On Oct 25, 2023, at 12:17 PM, John Forecast
<john(a)forecast.name <mailto:john@forecast.name>> wrote:
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
What does it mean for nodes to be displayed in that list? For example, 28NH isn't
connected by Ethernet (it is an endnode with a DDCMP link). So it isn't a neighbor
and that hardware address isn't meaningful.
paul
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
6:46 p.m.
On Oct 25, 2023, at 12:43 PM, John Forecast
<john(a)forecast.name> wrote:
So in theory, if there are multiple routers on the LAN, it would display the MAC address
of the router that forwarded the traffic for the node displayed, right? That's what
the next-hop cache (as opposed to the on-LAN cache) is.
paul
On Oct 25, 2023, at 12:34 PM, Paul Koning
<paulkoning(a)comcast.net> wrote:
That’s actually coming from the next-hop cache. The address should actually be labeled
“Next-Hop” and contain 41.250’s MAC address. It’s a leftover from the old implementation.
On Oct 25, 2023, at 12:17 PM, John Forecast
<john(a)forecast.name> wrote:
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
What does it mean for nodes to be displayed in that list? For example, 28NH isn't
connected by Ethernet (it is an endnode with a DDCMP link). So it isn't a neighbor
and that hardware address isn't meaningful.
paul
7:03 p.m.
On Oct 25, 2023, at 12:46 PM, Paul Koning
<paulkoning(a)comcast.net> wrote:
Yes, that’s correct. I just check and it’s actually dneigh that’s converting the
DECnet address into a MAC address. I need to change it to get the next hop from the
kernel.
John.
On Oct 25, 2023, at 12:43 PM, John Forecast
<john(a)forecast.name> wrote:
So in theory, if there are multiple routers on the LAN, it would display the MAC address
of the router that forwarded the traffic for the node displayed, right? That's what
the next-hop cache (as opposed to the on-LAN cache) is.
paul
On Oct 25, 2023, at 12:34 PM, Paul Koning
<paulkoning(a)comcast.net> wrote:
That’s actually coming from the next-hop cache. The address should actually be labeled
“Next-Hop” and contain 41.250’s MAC address. It’s a leftover from the old implementation.
On Oct 25, 2023, at 12:17 PM, John Forecast
<john(a)forecast.name> wrote:
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
What does it mean for nodes to be displayed in that list? For example, 28NH isn't
connected by Ethernet (it is an endnode with a DDCMP link). So it isn't a neighbor
and that hardware address isn't meaningful.
paul
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
7:43 p.m.
John,
Ok. I first thought MIM:: is 1.1::. I now mean 1.13::
systemctl status decnet3
? decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: active (exited) since Wed 2023-10-25 13:24:51 EDT; 1min 2s ago
Process: 1089 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1122 ExecStart=/usr/local/sbin/dnetd (code=exited, status=0/SUCCES>
Main PID: 1122 (code=exited, status=0/SUCCESS)
Tasks: 1 (limit: 37150)
Memory: 4.1M
CPU: 47ms
CGroup: /system.slice/decnet3.service
+-1133 /usr/local/sbin/dnetd
Oct 25 13:24:51 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 13:24:51 tsux1a systemd[1]: Finished decnet3.service - Load DECnet modul>
dneigh
Node HWtype HWaddress Flags MTU Iface
TSUX1A loop AA:00:04:00:9A:7C --- 65533 lo
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line enp4s0
I tried to add tsrtr to decnet.conf but dneigh showed that it routes to enp4s0 instead of
br0.
I tried to configure br0 instead of enp4s0 but it crashed during booting. systemctl
showed that loading failure with attempting connect to br0 interface.
It still can't see my pydecnet router (tsrtr 31.153).
Do you know how to configure decnet with br0 interface? How to set up a route to TSRTR
(31.153)?
________________________________
From: John Forecast <john(a)forecast.name>
Sent: Wednesday, October 25, 2023 12:17 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
On Oct 25, 2023, at 10:47 AM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
I successfully installed Linux DECnet (beta test) on Ubuntu 23.10 but have some problems
with that.
I executed 'sudo ./BuildAndInstall.sh' and followed instructions to enter, etc.
Also I started pydecnet (route to HECnet network) as bridge.
I tried to set host to my VAX simulator but got no route to host errors.
Also I tried to access 1.1:: but got no route host errors as well. Look that below.
dndir 1.1::
connect failed: No route to host
I see this as well.
sethost tsvmsb
sethost V1.0.4
Connecting to tsvmsb
connect (cterm): No route to host
ct_setup_link: error 0 connecting to host
I don’t have a VMS system set up to try this but I get further trying to connect to
RSX-11m+:
forecast@emulat:~$ sethost 41.200
sethost V1.0.4
Connecting to 41.200
ct_setup_link: error sending init sequence
sudo ncp show known nodes
Version mismatch
You don’t need the “sudo” for normal operation. The “Version mismatch” is in response to
the connection completing with the wrong
optional data. Did this system have the previous DECnet for Linux installed? If so, did
you follow the section in README.DECnet
about deleting old libraries.
sudo ncp copy known nodes from 1.1
Unable to connect to listener
Same problem as dndir above.
lsmod | grep decnet3
decnet3 86016 5
That looks good. What about the output from “systemctl status decnet3”
John.
Do you have any solutions with those problems? On my VAX simulator (TSVMSB), I was able
access HECnet without any problems.
I tried to set host to Ubuntu 23.10 but immediately got error - network is unreachable.
Tim
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Tuesday, October 24, 2023 10:50 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to
rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was
alway marked as
“experimental” and I never tried to get it running. If you need a DECnet
router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux
versions
were changes to the kernel framework APIs. By limiting the use of these
APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel
changes. So far,
I have had to make one additional conditional code change at kernel 6.5
and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal” is
still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management listener.
- There is now a subset implementation of “ncp” which is sufficient to support all
of the requests
available from the Linux network management listener. It does implement a “tell”
prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node
name database.
- The new kernel module now supports node counters as defined in the Network
Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi
releases. The current
procedure supports installation on systems derived from Debian or Fedora
although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
9:08 p.m.
Tim,
On Oct 25, 2023, at 1:43 PM, Timothy Stark
<fsword007(a)gmail.com> wrote:
John,
Ok. I first thought MIM:: is 1.1::. I now mean 1.13::
systemctl status decnet3
● decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: active (exited) since Wed 2023-10-25 13:24:51 EDT; 1min 2s ago
Process: 1089 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1122 ExecStart=/usr/local/sbin/dnetd (code=exited, status=0/SUCCES>
Main PID: 1122 (code=exited, status=0/SUCCESS)
Tasks: 1 (limit: 37150)
Memory: 4.1M
CPU: 47ms
CGroup: /system.slice/decnet3.service
└─1133 /usr/local/sbin/dnetd
Oct 25 13:24:51 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 13:24:51 tsux1a systemd[1]: Finished decnet3.service - Load DECnet modul>
That looks good.
dneigh
Node HWtype HWaddress Flags MTU Iface
TSUX1A loop AA:00:04:00:9A:7C --- 65533 lo
That also looks good, although there should be a second line with the designated router.
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line enp4s0
OK.
I tried to add tsrtr to decnet.conf but dneigh showed
that it routes to enp4s0 instead of br0.
This release is an end-node only implementation. There can only be a single “executor”
line which defines the local name, address and
physical (ethernet or wifi) device. All other lines define node address/name pairs. There
is no concept of setting a route to a node like in TCP/IP.
TSUX1A is expected to learn about TSRTR via a multicast message on the LAN connecting
TSRTR and TSUX1A.
I tried to configure br0 instead of enp4s0 but it
crashed during booting. systemctl showed that loading failure with attempting connect to
br0 interface.
Yes, that seems reasonable.
It still can't see my pydecnet router (tsrtr
31.153).
Do you know how to configure decnet with br0 interface? How to set up a route to TSRTR
(31.153)?
I think the main problem here is that TSUX1A is not seeing TSRTR. How are they connected?
What does the WEB interface into TSRTR show?
John.
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 12:17 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
On Oct 25, 2023, at 10:47 AM, Timothy Stark
<fsword007(a)gmail.com <mailto:fsword007@gmail.com>> wrote:
I successfully installed Linux DECnet (beta test) on Ubuntu 23.10 but have some problems
with that.
I executed 'sudo ./BuildAndInstall.sh' and followed instructions to enter, etc.
Also I started pydecnet (route to HECnet network) as bridge.
I tried to set host to my VAX simulator but got no route to host errors.
Also I tried to access 1.1:: but got no route host errors as well. Look that below.
dndir 1.1::
connect failed: No route to host
I see this as well.
sethost tsvmsb
sethost V1.0.4
Connecting to tsvmsb
connect (cterm): No route to host
ct_setup_link: error 0 connecting to host
I don’t have a VMS system set up to try this but I get further trying to connect to
RSX-11m+:
forecast@emulat:~$ sethost 41.200
sethost V1.0.4
Connecting to 41.200
ct_setup_link: error sending init sequence
sudo ncp show known nodes
Version mismatch
You don’t need the “sudo” for normal operation. The “Version mismatch” is in response to
the connection completing with the wrong
optional data. Did this system have the previous DECnet for Linux installed? If so, did
you follow the section in README.DECnet
about deleting old libraries.
sudo ncp copy known nodes from 1.1
Unable to connect to listener
Same problem as dndir above.
lsmod | grep decnet3
decnet3 86016 5
That looks good. What about the output from “systemctl status decnet3”
John.
Do you have any solutions with those problems?
On my VAX simulator (TSVMSB), I was able access HECnet without any problems.
I tried to set host to Ubuntu 23.10 but immediately got error - network is unreachable.
Tim
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Tuesday, October 24, 2023 10:50 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet
<https://github.com/JohnForecast/RaspbianDECnet>> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to
rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was
alway marked as
“experimental” and I never tried to get it running. If you need a DECnet
router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux
versions
were changes to the kernel framework APIs. By limiting the use of these
APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel
changes. So far,
I have had to make one additional conditional code change at kernel 6.5
and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal”
is still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management
listener.
- There is now a subset implementation of “ncp” which is sufficient to support
all of the requests
available from the Linux network management listener. It does implement a
“tell” prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node
name database.
- The new kernel module now supports node counters as defined in the Network
Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi
releases. The current
procedure supports installation on systems derived from Debian or Fedora
although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet
<https://github.com/JohnForecast/LinuxDECnet>>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
9:41 p.m.
Here is my tsrtr.conf for pydecnet
circuit mul-0 Multinet <hecnet gw address>:connect
circuit tap-0 Ethernet tap:tap0
routing 31.153 --type l1router --bct1=20
nsp --qmax=5
node 31.153 TSRTR
node @nodenames.dat
system --ident "TSRTR PyDECnet router (Ubuntu system)"
I did not enable web interface into TSRTR server. I need to add circuit to connect an
end-node decnet for TSUX1A.
I have br0 and enp4s0 interface. Which interface should be added so that TSUX1A can see
TSRTR router?
I have tap-0 line so that VAX emulators can see HECnet world.
Tim
________________________________
From: John Forecast <john(a)forecast.name>
Sent: Wednesday, October 25, 2023 3:08 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
On Oct 25, 2023, at 1:43 PM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
John,
Ok. I first thought MIM:: is 1.1::. I now mean 1.13::
systemctl status decnet3
● decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: active (exited) since Wed 2023-10-25 13:24:51 EDT; 1min 2s ago
Process: 1089 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1122 ExecStart=/usr/local/sbin/dnetd (code=exited, status=0/SUCCES>
Main PID: 1122 (code=exited, status=0/SUCCESS)
Tasks: 1 (limit: 37150)
Memory: 4.1M
CPU: 47ms
CGroup: /system.slice/decnet3.service
└─1133 /usr/local/sbin/dnetd
Oct 25 13:24:51 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 13:24:51 tsux1a systemd[1]: Finished decnet3.service - Load DECnet modul>
That looks good.
dneigh
Node HWtype HWaddress Flags MTU Iface
TSUX1A loop AA:00:04:00:9A:7C --- 65533 lo
That also looks good, although there should be a second line with the designated router.
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line enp4s0
OK.
I tried to add tsrtr to decnet.conf but dneigh showed that it routes to enp4s0 instead of
br0.
This release is an end-node only implementation. There can only be a single “executor”
line which defines the local name, address and
physical (ethernet or wifi) device. All other lines define node address/name pairs. There
is no concept of setting a route to a node like in TCP/IP.
TSUX1A is expected to learn about TSRTR via a multicast message on the LAN connecting
TSRTR and TSUX1A.
I tried to configure br0 instead of enp4s0 but it crashed during booting. systemctl
showed that loading failure with attempting connect to br0 interface.
Yes, that seems reasonable.
It still can't see my pydecnet router (tsrtr 31.153).
Do you know how to configure decnet with br0 interface? How to set up a route to TSRTR
(31.153)?
I think the main problem here is that TSUX1A is not seeing TSRTR. How are they connected?
What does the WEB interface into TSRTR show?
John.
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 12:17 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
On Oct 25, 2023, at 10:47 AM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
I successfully installed Linux DECnet (beta test) on Ubuntu 23.10 but have some problems
with that.
I executed 'sudo ./BuildAndInstall.sh' and followed instructions to enter, etc.
Also I started pydecnet (route to HECnet network) as bridge.
I tried to set host to my VAX simulator but got no route to host errors.
Also I tried to access 1.1:: but got no route host errors as well. Look that below.
dndir 1.1::
connect failed: No route to host
I see this as well.
sethost tsvmsb
sethost V1.0.4
Connecting to tsvmsb
connect (cterm): No route to host
ct_setup_link: error 0 connecting to host
I don’t have a VMS system set up to try this but I get further trying to connect to
RSX-11m+:
forecast@emulat:~$ sethost 41.200
sethost V1.0.4
Connecting to 41.200
ct_setup_link: error sending init sequence
sudo ncp show known nodes
Version mismatch
You don’t need the “sudo” for normal operation. The “Version mismatch” is in response to
the connection completing with the wrong
optional data. Did this system have the previous DECnet for Linux installed? If so, did
you follow the section in README.DECnet
about deleting old libraries.
sudo ncp copy known nodes from 1.1
Unable to connect to listener
Same problem as dndir above.
lsmod | grep decnet3
decnet3 86016 5
That looks good. What about the output from “systemctl status decnet3”
John.
Do you have any solutions with those problems? On my VAX simulator (TSVMSB), I was able
access HECnet without any problems.
I tried to set host to Ubuntu 23.10 but immediately got error - network is unreachable.
Tim
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Tuesday, October 24, 2023 10:50 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to
rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was
alway marked as
“experimental” and I never tried to get it running. If you need a DECnet
router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux
versions
were changes to the kernel framework APIs. By limiting the use of these
APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel
changes. So far,
I have had to make one additional conditional code change at kernel 6.5
and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal” is
still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management listener.
- There is now a subset implementation of “ncp” which is sufficient to support all
of the requests
available from the Linux network management listener. It does implement a “tell”
prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node
name database.
- The new kernel module now supports node counters as defined in the Network
Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi
releases. The current
procedure supports installation on systems derived from Debian or Fedora
although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
11:05 p.m.
Tim,
What does your physical topology look like? You have TSRTR with routing between an
incoming Multinet connection and tap0, which is a
software construct so your VAX emulators can connect to HECnet. What is TSUX1A and how is
it connected to TSRTR? The output of “ifconfig”
or “ip addr” would probably be useful here - executed on both TSRTR and TSUX1A.
John.
On Oct 25, 2023, at 3:41 PM, Timothy Stark
<fsword007(a)gmail.com> wrote:
Here is my tsrtr.conf for pydecnet
circuit mul-0 Multinet <hecnet gw address>:connect
circuit tap-0 Ethernet tap:tap0
routing 31.153 --type l1router --bct1=20
nsp --qmax=5
node 31.153 TSRTR
node @nodenames.dat
system --ident "TSRTR PyDECnet router (Ubuntu system)"
I did not enable web interface into TSRTR server. I need to add circuit to connect an
end-node decnet for TSUX1A.
I have br0 and enp4s0 interface. Which interface should be added so that TSUX1A can see
TSRTR router?
I have tap-0 line so that VAX emulators can see HECnet world.
Tim
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 3:08 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
On Oct 25, 2023, at 1:43 PM, Timothy Stark
<fsword007(a)gmail.com <mailto:fsword007@gmail.com>> wrote:
John,
Ok. I first thought MIM:: is 1.1::. I now mean 1.13::
systemctl status decnet3
● decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: active (exited) since Wed 2023-10-25 13:24:51 EDT; 1min 2s ago
Process: 1089 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1122 ExecStart=/usr/local/sbin/dnetd (code=exited, status=0/SUCCES>
Main PID: 1122 (code=exited, status=0/SUCCESS)
Tasks: 1 (limit: 37150)
Memory: 4.1M
CPU: 47ms
CGroup: /system.slice/decnet3.service
└─1133 /usr/local/sbin/dnetd
Oct 25 13:24:51 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 13:24:51 tsux1a systemd[1]: Finished decnet3.service - Load DECnet modul>
That looks good.
dneigh
Node HWtype HWaddress Flags MTU Iface
TSUX1A loop AA:00:04:00:9A:7C --- 65533 lo
That also looks good, although there should be a second line with the designated router.
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line enp4s0
OK.
I tried to add tsrtr to decnet.conf but dneigh
showed that it routes to enp4s0 instead of br0.
This release is an end-node only implementation. There can only be a single “executor”
line which defines the local name, address and
physical (ethernet or wifi) device. All other lines define node address/name pairs. There
is no concept of setting a route to a node like in TCP/IP.
TSUX1A is expected to learn about TSRTR via a multicast message on the LAN connecting
TSRTR and TSUX1A.
I tried to configure br0 instead of enp4s0 but it
crashed during booting. systemctl showed that loading failure with attempting connect to
br0 interface.
Yes, that seems reasonable.
It still can't see my pydecnet router (tsrtr
31.153).
Do you know how to configure decnet with br0 interface? How to set up a route to TSRTR
(31.153)?
I think the main problem here is that TSUX1A is not seeing TSRTR. How are they connected?
What does the WEB interface into TSRTR show?
John.
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 12:17 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se> On Oct 25, 2023, at 10:47 AM, Timothy Stark
<fsword007(a)gmail.com <mailto:fsword007@gmail.com>> wrote:
I successfully installed Linux DECnet (beta test) on Ubuntu 23.10 but have some problems
with that.
I executed 'sudo ./BuildAndInstall.sh' and followed instructions to enter, etc.
Also I started pydecnet (route to HECnet network) as bridge.
I tried to set host to my VAX simulator but got no route to host errors.
Also I tried to access 1.1:: but got no route host errors as well. Look that below.
dndir 1.1::
connect failed: No route to host
I see this as well.
sethost tsvmsb
sethost V1.0.4
Connecting to tsvmsb
connect (cterm): No route to host
ct_setup_link: error 0 connecting to host
I don’t have a VMS system set up to try this but I get further trying to connect to
RSX-11m+:
forecast@emulat:~$ sethost 41.200
sethost V1.0.4
Connecting to 41.200
ct_setup_link: error sending init sequence
sudo ncp show known nodes
Version mismatch
You don’t need the “sudo” for normal operation. The “Version mismatch” is in response to
the connection completing with the wrong
optional data. Did this system have the previous DECnet for Linux installed? If so, did
you follow the section in README.DECnet
about deleting old libraries.
sudo ncp copy known nodes from 1.1
Unable to connect to listener
Same problem as dndir above.
lsmod | grep decnet3
decnet3 86016 5
That looks good. What about the output from “systemctl status decnet3”
John.
Do you have any solutions with those problems?
On my VAX simulator (TSVMSB), I was able access HECnet without any problems.
I tried to set host to Ubuntu 23.10 but immediately got error - network is unreachable.
Tim
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Tuesday, October 24, 2023 10:50 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet
<https://github.com/JohnForecast/RaspbianDECnet>> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to
rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was
alway marked as
“experimental” and I never tried to get it running. If you need a DECnet
router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux
versions
were changes to the kernel framework APIs. By limiting the use of these
APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel
changes. So far,
I have had to make one additional conditional code change at kernel 6.5
and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal”
is still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management
listener.
- There is now a subset implementation of “ncp” which is sufficient to support
all of the requests
available from the Linux network management listener. It does implement a
“tell” prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node
name database.
- The new kernel module now supports node counters as defined in the Network
Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi
releases. The current
procedure supports installation on systems derived from Debian or Fedora
although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet
<https://github.com/JohnForecast/LinuxDECnet>>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
11:29 p.m.
John,
I set virtual br0 for 8 tap devices that VAX emulators and others use. br0 is the main
routing. virbr0 is for QEMU VM (OpenVMS x86). TSUX1A is IP 192.168.1.228.
TSUX1A is Ubuntu 23.10 system on my embedded box. TSRTR is pydecnet router for HECnet
world access that connected to tap0. I am figuring out why TSRTR and TSUX1A can't see
each other. I am only able access TSVMSB (on SIMH VAX emulator) that uses tap1 through
TCP/IP but not DECnet.
TSUX1A is connected to my Verizon FiOS router for Internet access.
Here is ifconfig output below.
Tim
br0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 192.168.1.228 netmask 255.255.255.0 broadcast 192.168.1.255
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c prefixlen 64 scopeid
0x0<global>
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39 prefixlen 64 scopeid
0x0<global>
inet6 fe80::c288:756:b4bc:213e prefixlen 64 scopeid 0x20<link>
ether 3a:22:1a:4c:ea:5a txqueuelen 1000 (Ethernet)
RX packets 85254 bytes 10370418 (10.3 MB)
RX errors 0 dropped 8364 overruns 0 frame 0
TX packets 8096 bytes 813053 (813.0 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
enp4s0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
ether aa:00:04:00:9a:7c txqueuelen 1000 (Ethernet)
RX packets 100400 bytes 16878747 (16.8 MB)
RX errors 0 dropped 14 overruns 0 frame 0
TX packets 21289 bytes 2748243 (2.7 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
lo: flags=73<UP,LOOPBACK,RUNNING> mtu 65536
inet 127.0.0.1 netmask 255.0.0.0
inet6 ::1 prefixlen 128 scopeid 0x10<host>
loop txqueuelen 1000 (Local Loopback)
RX packets 1978 bytes 538618 (538.6 KB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 1978 bytes 538618 (538.6 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet6 fe80::884d:baff:fe98:9c38 prefixlen 64 scopeid 0x20<link>
ether 8a:4d:ba:98:9c:38 txqueuelen 1000 (Ethernet)
RX packets 5181 bytes 1525244 (1.5 MB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 74004 bytes 7346087 (7.3 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap1: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 6a:29:52:18:5a:14 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap2: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:90:0b:15:c2:7c txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap3: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 86:ae:14:ac:75:27 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap4: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether e6:d7:b6:a4:6e:2b txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap5: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 12:2f:e4:99:a2:70 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap6: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:21:92:ad:88:a0 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap7: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 9e:fd:b7:cd:a3:e6 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
virbr0: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
inet 192.168.122.1 netmask 255.255.255.0 broadcast 192.168.122.255
ether 52:54:00:76:61:73 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
ip addr:
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default
qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: enp4s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether aa:00:04:00:9a:7c brd ff:ff:ff:ff:ff:ff permaddr 00:c0:08:97:1a:2a
3: wlp3s0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen
1000
link/ether 54:14:f3:52:56:9d brd ff:ff:ff:ff:ff:ff
4: br0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group
default qlen 1000
link/ether 3a:22:1a:4c:ea:5a brd ff:ff:ff:ff:ff:ff
inet 192.168.1.228/24 brd 192.168.1.255 scope global dynamic noprefixroute br0
valid_lft 51908sec preferred_lft 51908sec
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c/64 scope global temporary dynamic
valid_lft 7193sec preferred_lft 7193sec
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39/64 scope global dynamic mngtmpaddr
noprefixroute
valid_lft 7193sec preferred_lft 7193sec
inet6 fe80::c288:756:b4bc:213e/64 scope link noprefixroute
valid_lft forever preferred_lft forever
5: virbr0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN
group default qlen 1000
link/ether 52:54:00:76:61:73 brd ff:ff:ff:ff:ff:ff
inet 192.168.122.1/24 brd 192.168.122.255 scope global virbr0
valid_lft forever preferred_lft forever
6: tap0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether 8a:4d:ba:98:9c:38 brd ff:ff:ff:ff:ff:ff
inet6 fe80::884d:baff:fe98:9c38/64 scope link
valid_lft forever preferred_lft forever
7: tap1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 6a:29:52:18:5a:14 brd ff:ff:ff:ff:ff:ff
8: tap2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:90:0b:15:c2:7c brd ff:ff:ff:ff:ff:ff
9: tap3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 86:ae:14:ac:75:27 brd ff:ff:ff:ff:ff:ff
10: tap4: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether e6:d7:b6:a4:6e:2b brd ff:ff:ff:ff:ff:ff
11: tap5: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 12:2f:e4:99:a2:70 brd ff:ff:ff:ff:ff:ff
12: tap6: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:21:92:ad:88:a0 brd ff:ff:ff:ff:ff:ff
13: tap7: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 9e:fd:b7:cd:a3:e6 brd ff:ff:ff:ff:ff:ff
________________________________
From: John Forecast <john(a)forecast.name>
Sent: Wednesday, October 25, 2023 5:05 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
What does your physical topology look like? You have TSRTR with routing between an
incoming Multinet connection and tap0, which is a
software construct so your VAX emulators can connect to HECnet. What is TSUX1A and how is
it connected to TSRTR? The output of “ifconfig”
or “ip addr” would probably be useful here - executed on both TSRTR and TSUX1A.
John.
On Oct 25, 2023, at 3:41 PM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
Here is my tsrtr.conf for pydecnet
circuit mul-0 Multinet <hecnet gw address>:connect
circuit tap-0 Ethernet tap:tap0
routing 31.153 --type l1router --bct1=20
nsp --qmax=5
node 31.153 TSRTR
node @nodenames.dat
system --ident "TSRTR PyDECnet router (Ubuntu system)"
I did not enable web interface into TSRTR server. I need to add circuit to connect an
end-node decnet for TSUX1A.
I have br0 and enp4s0 interface. Which interface should be added so that TSUX1A can see
TSRTR router?
I have tap-0 line so that VAX emulators can see HECnet world.
Tim
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 3:08 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
On Oct 25, 2023, at 1:43 PM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
John,
Ok. I first thought MIM:: is 1.1::. I now mean 1.13::
systemctl status decnet3
● decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: active (exited) since Wed 2023-10-25 13:24:51 EDT; 1min 2s ago
Process: 1089 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1122 ExecStart=/usr/local/sbin/dnetd (code=exited, status=0/SUCCES>
Main PID: 1122 (code=exited, status=0/SUCCESS)
Tasks: 1 (limit: 37150)
Memory: 4.1M
CPU: 47ms
CGroup: /system.slice/decnet3.service
└─1133 /usr/local/sbin/dnetd
Oct 25 13:24:51 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 13:24:51 tsux1a systemd[1]: Finished decnet3.service - Load DECnet modul>
That looks good.
dneigh
Node HWtype HWaddress Flags MTU Iface
TSUX1A loop AA:00:04:00:9A:7C --- 65533 lo
That also looks good, although there should be a second line with the designated router.
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line enp4s0
OK.
I tried to add tsrtr to decnet.conf but dneigh showed that it routes to enp4s0 instead of
br0.
This release is an end-node only implementation. There can only be a single “executor”
line which defines the local name, address and
physical (ethernet or wifi) device. All other lines define node address/name pairs. There
is no concept of setting a route to a node like in TCP/IP.
TSUX1A is expected to learn about TSRTR via a multicast message on the LAN connecting
TSRTR and TSUX1A.
I tried to configure br0 instead of enp4s0 but it crashed during booting. systemctl
showed that loading failure with attempting connect to br0 interface.
Yes, that seems reasonable.
It still can't see my pydecnet router (tsrtr 31.153).
Do you know how to configure decnet with br0 interface? How to set up a route to TSRTR
(31.153)?
I think the main problem here is that TSUX1A is not seeing TSRTR. How are they connected?
What does the WEB interface into TSRTR show?
John.
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 12:17 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
On Oct 25, 2023, at 10:47 AM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
I successfully installed Linux DECnet (beta test) on Ubuntu 23.10 but have some problems
with that.
I executed 'sudo ./BuildAndInstall.sh' and followed instructions to enter, etc.
Also I started pydecnet (route to HECnet network) as bridge.
I tried to set host to my VAX simulator but got no route to host errors.
Also I tried to access 1.1:: but got no route host errors as well. Look that below.
dndir 1.1::
connect failed: No route to host
I see this as well.
sethost tsvmsb
sethost V1.0.4
Connecting to tsvmsb
connect (cterm): No route to host
ct_setup_link: error 0 connecting to host
I don’t have a VMS system set up to try this but I get further trying to connect to
RSX-11m+:
forecast@emulat:~$ sethost 41.200
sethost V1.0.4
Connecting to 41.200
ct_setup_link: error sending init sequence
sudo ncp show known nodes
Version mismatch
You don’t need the “sudo” for normal operation. The “Version mismatch” is in response to
the connection completing with the wrong
optional data. Did this system have the previous DECnet for Linux installed? If so, did
you follow the section in README.DECnet
about deleting old libraries.
sudo ncp copy known nodes from 1.1
Unable to connect to listener
Same problem as dndir above.
lsmod | grep decnet3
decnet3 86016 5
That looks good. What about the output from “systemctl status decnet3”
John.
Do you have any solutions with those problems? On my VAX simulator (TSVMSB), I was able
access HECnet without any problems.
I tried to set host to Ubuntu 23.10 but immediately got error - network is unreachable.
Tim
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Tuesday, October 24, 2023 10:50 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to
rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was
alway marked as
“experimental” and I never tried to get it running. If you need a DECnet
router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux
versions
were changes to the kernel framework APIs. By limiting the use of these
APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel
changes. So far,
I have had to make one additional conditional code change at kernel 6.5
and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal” is
still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management listener.
- There is now a subset implementation of “ncp” which is sufficient to support all
of the requests
available from the Linux network management listener. It does implement a “tell”
prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node
name database.
- The new kernel module now supports node counters as defined in the Network
Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi
releases. The current
procedure supports installation on systems derived from Debian or Fedora
although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
11:53 p.m.
Tim,
I’m not familiar with the FIOS router configuration. I’m going to guess that there’s
one or more ethernet’s coming out of it; one going to TSRTR
and another going to TSUX1A. On TSRTR that is probably “enp4s0”. You will need to add a
circuit to pydecnet so it does routing on your
ethernet segment, something like:
circuit eth-0 Ethernet pcap:enp4s0
John.
On Oct 25, 2023, at 5:29 PM, Timothy Stark
<fsword007(a)gmail.com> wrote:
John,
I set virtual br0 for 8 tap devices that VAX emulators and others use. br0 is the main
routing. virbr0 is for QEMU VM (OpenVMS x86). TSUX1A is IP 192.168.1.228.
TSUX1A is Ubuntu 23.10 system on my embedded box. TSRTR is pydecnet router for HECnet
world access that connected to tap0. I am figuring out why TSRTR and TSUX1A can't see
each other. I am only able access TSVMSB (on SIMH VAX emulator) that uses tap1 through
TCP/IP but not DECnet.
TSUX1A is connected to my Verizon FiOS router for Internet access.
Here is ifconfig output below.
Tim
br0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 192.168.1.228 netmask 255.255.255.0 broadcast 192.168.1.255
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c prefixlen 64 scopeid
0x0<global>
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39 prefixlen 64 scopeid
0x0<global>
inet6 fe80::c288:756:b4bc:213e prefixlen 64 scopeid 0x20<link>
ether 3a:22:1a:4c:ea:5a txqueuelen 1000 (Ethernet)
RX packets 85254 bytes 10370418 (10.3 MB)
RX errors 0 dropped 8364 overruns 0 frame 0
TX packets 8096 bytes 813053 (813.0 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
enp4s0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
ether aa:00:04:00:9a:7c txqueuelen 1000 (Ethernet)
RX packets 100400 bytes 16878747 (16.8 MB)
RX errors 0 dropped 14 overruns 0 frame 0
TX packets 21289 bytes 2748243 (2.7 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
lo: flags=73<UP,LOOPBACK,RUNNING> mtu 65536
inet 127.0.0.1 netmask 255.0.0.0
inet6 ::1 prefixlen 128 scopeid 0x10<host>
loop txqueuelen 1000 (Local Loopback)
RX packets 1978 bytes 538618 (538.6 KB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 1978 bytes 538618 (538.6 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet6 fe80::884d:baff:fe98:9c38 prefixlen 64 scopeid 0x20<link>
ether 8a:4d:ba:98:9c:38 txqueuelen 1000 (Ethernet)
RX packets 5181 bytes 1525244 (1.5 MB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 74004 bytes 7346087 (7.3 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap1: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 6a:29:52:18:5a:14 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap2: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:90:0b:15:c2:7c txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap3: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 86:ae:14:ac:75:27 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap4: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether e6:d7:b6:a4:6e:2b txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap5: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 12:2f:e4:99:a2:70 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap6: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:21:92:ad:88:a0 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap7: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 9e:fd:b7:cd:a3:e6 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
virbr0: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
inet 192.168.122.1 netmask 255.255.255.0 broadcast 192.168.122.255
ether 52:54:00:76:61:73 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
ip addr:
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default
qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: enp4s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether aa:00:04:00:9a:7c brd ff:ff:ff:ff:ff:ff permaddr 00:c0:08:97:1a:2a
3: wlp3s0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen
1000
link/ether 54:14:f3:52:56:9d brd ff:ff:ff:ff:ff:ff
4: br0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group
default qlen 1000
link/ether 3a:22:1a:4c:ea:5a brd ff:ff:ff:ff:ff:ff
inet 192.168.1.228/24 brd 192.168.1.255 scope global dynamic noprefixroute br0
valid_lft 51908sec preferred_lft 51908sec
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c/64 scope global temporary dynamic
valid_lft 7193sec preferred_lft 7193sec
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39/64 scope global dynamic mngtmpaddr
noprefixroute
valid_lft 7193sec preferred_lft 7193sec
inet6 fe80::c288:756:b4bc:213e/64 scope link noprefixroute
valid_lft forever preferred_lft forever
5: virbr0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN
group default qlen 1000
link/ether 52:54:00:76:61:73 brd ff:ff:ff:ff:ff:ff
inet 192.168.122.1/24 brd 192.168.122.255 scope global virbr0
valid_lft forever preferred_lft forever
6: tap0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether 8a:4d:ba:98:9c:38 brd ff:ff:ff:ff:ff:ff
inet6 fe80::884d:baff:fe98:9c38/64 scope link
valid_lft forever preferred_lft forever
7: tap1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 6a:29:52:18:5a:14 brd ff:ff:ff:ff:ff:ff
8: tap2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:90:0b:15:c2:7c brd ff:ff:ff:ff:ff:ff
9: tap3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 86:ae:14:ac:75:27 brd ff:ff:ff:ff:ff:ff
10: tap4: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether e6:d7:b6:a4:6e:2b brd ff:ff:ff:ff:ff:ff
11: tap5: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 12:2f:e4:99:a2:70 brd ff:ff:ff:ff:ff:ff
12: tap6: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:21:92:ad:88:a0 brd ff:ff:ff:ff:ff:ff
13: tap7: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 9e:fd:b7:cd:a3:e6 brd ff:ff:ff:ff:ff:ff
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 5:05 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
What does your physical topology look like? You have TSRTR with routing between an
incoming Multinet connection and tap0, which is a
software construct so your VAX emulators can connect to HECnet. What is TSUX1A and how is
it connected to TSRTR? The output of “ifconfig”
or “ip addr” would probably be useful here - executed on both TSRTR and TSUX1A.
John.
On Oct 25, 2023, at 3:41 PM, Timothy Stark <fsword007(a)gmail.com
<mailto:fsword007@gmail.com>> wrote:
Here is my tsrtr.conf for pydecnet
circuit mul-0 Multinet <hecnet gw address>:connect
circuit tap-0 Ethernet tap:tap0
routing 31.153 --type l1router --bct1=20
nsp --qmax=5
node 31.153 TSRTR
node @nodenames.dat
system --ident "TSRTR PyDECnet router (Ubuntu system)"
I did not enable web interface into TSRTR server. I need to add circuit to connect an
end-node decnet for TSUX1A.
I have br0 and enp4s0 interface. Which interface should be added so that TSUX1A can see
TSRTR router?
I have tap-0 line so that VAX emulators can see HECnet world.
Tim
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 3:08 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
On Oct 25, 2023, at 1:43 PM, Timothy Stark <fsword007(a)gmail.com
<mailto:fsword007@gmail.com>> wrote:
John,
Ok. I first thought MIM:: is 1.1::. I now mean 1.13::
systemctl status decnet3
● decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: active (exited) since Wed 2023-10-25 13:24:51 EDT; 1min 2s ago
Process: 1089 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1122 ExecStart=/usr/local/sbin/dnetd (code=exited, status=0/SUCCES>
Main PID: 1122 (code=exited, status=0/SUCCESS)
Tasks: 1 (limit: 37150)
Memory: 4.1M
CPU: 47ms
CGroup: /system.slice/decnet3.service
└─1133 /usr/local/sbin/dnetd
Oct 25 13:24:51 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 13:24:51 tsux1a systemd[1]: Finished decnet3.service - Load DECnet modul>
That looks good.
dneigh
Node HWtype HWaddress Flags MTU Iface
TSUX1A loop AA:00:04:00:9A:7C --- 65533 lo
That also looks good, although there should be a second line with the designated router.
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line enp4s0
OK.
I tried to add tsrtr to decnet.conf but dneigh showed that it routes to enp4s0 instead of
br0.
This release is an end-node only implementation. There can only be a single “executor”
line which defines the local name, address and
physical (ethernet or wifi) device. All other lines define node address/name pairs. There
is no concept of setting a route to a node like in TCP/IP.
TSUX1A is expected to learn about TSRTR via a multicast message on the LAN connecting
TSRTR and TSUX1A.
I tried to configure br0 instead of enp4s0 but it crashed during booting. systemctl
showed that loading failure with attempting connect to br0 interface.
Yes, that seems reasonable.
It still can't see my pydecnet router (tsrtr 31.153).
Do you know how to configure decnet with br0 interface? How to set up a route to TSRTR
(31.153)?
I think the main problem here is that TSUX1A is not seeing TSRTR. How are they connected?
What does the WEB interface into TSRTR show?
John.
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 12:17 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
On Oct 25, 2023, at 10:47 AM, Timothy Stark <fsword007(a)gmail.com
<mailto:fsword007@gmail.com>> wrote:
I successfully installed Linux DECnet (beta test) on Ubuntu 23.10 but have some problems
with that.
I executed 'sudo ./BuildAndInstall.sh' and followed instructions to enter, etc.
Also I started pydecnet (route to HECnet network) as bridge.
I tried to set host to my VAX simulator but got no route to host errors.
Also I tried to access 1.1:: but got no route host errors as well. Look that below.
dndir 1.1::
connect failed: No route to host
I see this as well.
sethost tsvmsb
sethost V1.0.4
Connecting to tsvmsb
connect (cterm): No route to host
ct_setup_link: error 0 connecting to host
I don’t have a VMS system set up to try this but I get further trying to connect to
RSX-11m+:
forecast@emulat:~$ sethost 41.200
sethost V1.0.4
Connecting to 41.200
ct_setup_link: error sending init sequence
sudo ncp show known nodes
Version mismatch
You don’t need the “sudo” for normal operation. The “Version mismatch” is in response to
the connection completing with the wrong
optional data. Did this system have the previous DECnet for Linux installed? If so, did
you follow the section in README.DECnet
about deleting old libraries.
sudo ncp copy known nodes from 1.1
Unable to connect to listener
Same problem as dndir above.
lsmod | grep decnet3
decnet3 86016 5
That looks good. What about the output from “systemctl status decnet3”
John.
Do you have any solutions with those problems? On my VAX simulator (TSVMSB), I was able
access HECnet without any problems.
I tried to set host to Ubuntu 23.10 but immediately got error - network is unreachable.
Tim
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Tuesday, October 24, 2023 10:50 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet
<https://github.com/JohnForecast/RaspbianDECnet>> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to
rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was
alway marked as
“experimental” and I never tried to get it running. If you need a DECnet
router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux
versions
were changes to the kernel framework APIs. By limiting the use of these
APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel
changes. So far,
I have had to make one additional conditional code change at kernel 6.5
and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal”
is still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management
listener.
- There is now a subset implementation of “ncp” which is sufficient to support
all of the requests
available from the Linux network management listener. It does implement a
“tell” prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node
name database.
- The new kernel module now supports node counters as defined in the Network
Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi
releases. The current
procedure supports installation on systems derived from Debian or Fedora
although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet
<https://github.com/JohnForecast/LinuxDECnet>>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
26 Oct
26 Oct
12:59 a.m.
No, it did not work. TSRTR and VAX emulators can't see HECnet world with enp4s0. I
tried with br0 and it worked fine. I do not need tap0 anymore.
I tried to change to br0 by re-build configuration files with Linux DECnet. decnet3
crashed when it attempted to connect br0 interface. It said node_address is missing.
I created br0 with using brctl commands. When I upgraded Ubuntu systems for some years,
it preserved br0 configurations.
Tim
I changed a line in tsrtr.conf from pcap:tap0 line:
circuit eth-0 Ethernet pcap:br0
$ systemctl status decnet3
× decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: failed (Result: exit-code) since Wed 2023-10-25 18:29:05 EDT; 58s >
Process: 1097 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1140 ExecStart=/usr/local/sbin/dnetd (code=exited, status=255/EXCE>
Main PID: 1140 (code=exited, status=255/EXCEPTION)
CPU: 43ms
Oct 25 18:29:05 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 18:29:05 tsux1a dnetd[1140]: /proc/sys/net/decnet/node_address missing, >
Oct 25 18:29:05 tsux1a systemd[1]: decnet3.service: Main process exited, code=e>
Oct 25 18:29:05 tsux1a systemd[1]: decnet3.service: Failed with result 'exit-co>
Oct 25 18:29:05 tsux1a systemd[1]: Failed to start decnet3.service - Load DECne>
$ dneigh
Error: can not open DECnet neigh file: /proc/net/decnet_neigh: No such file or directory
$ brctl show
bridge name bridge id STP enabled interfaces
br0 8000.3a221a4cea5a yes enp4s0
virbr0 8000.525400766173 yes
$ cat /proc/net/decnet_dev
Name Flags T1 Timer1 T3 Timer3 T4 Timer4 BlkSize Pri State DevType Router
Peer
lo B 0000 0000 0000 0000 0000 0000 65533 000 00 loopback
br0 B 0000 0000 0000 0010 0000 0000 1498 000 00 ethernet
________________________________
From: John Forecast <john(a)forecast.name>
Sent: Wednesday, October 25, 2023 5:53 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
I’m not familiar with the FIOS router configuration. I’m going to guess that there’s
one or more ethernet’s coming out of it; one going to TSRTR
and another going to TSUX1A. On TSRTR that is probably “enp4s0”. You will need to add a
circuit to pydecnet so it does routing on your
ethernet segment, something like:
circuit eth-0 Ethernet pcap:enp4s0
John.
On Oct 25, 2023, at 5:29 PM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
John,
I set virtual br0 for 8 tap devices that VAX emulators and others use. br0 is the main
routing. virbr0 is for QEMU VM (OpenVMS x86). TSUX1A is IP 192.168.1.228.
TSUX1A is Ubuntu 23.10 system on my embedded box. TSRTR is pydecnet router for HECnet
world access that connected to tap0. I am figuring out why TSRTR and TSUX1A can't see
each other. I am only able access TSVMSB (on SIMH VAX emulator) that uses tap1 through
TCP/IP but not DECnet.
TSUX1A is connected to my Verizon FiOS router for Internet access.
Here is ifconfig output below.
Tim
br0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 192.168.1.228 netmask 255.255.255.0 broadcast 192.168.1.255
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c prefixlen 64 scopeid
0x0<global>
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39 prefixlen 64 scopeid
0x0<global>
inet6 fe80::c288:756:b4bc:213e prefixlen 64 scopeid 0x20<link>
ether 3a:22:1a:4c:ea:5a txqueuelen 1000 (Ethernet)
RX packets 85254 bytes 10370418 (10.3 MB)
RX errors 0 dropped 8364 overruns 0 frame 0
TX packets 8096 bytes 813053 (813.0 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
enp4s0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
ether aa:00:04:00:9a:7c txqueuelen 1000 (Ethernet)
RX packets 100400 bytes 16878747 (16.8 MB)
RX errors 0 dropped 14 overruns 0 frame 0
TX packets 21289 bytes 2748243 (2.7 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
lo: flags=73<UP,LOOPBACK,RUNNING> mtu 65536
inet 127.0.0.1 netmask 255.0.0.0
inet6 ::1 prefixlen 128 scopeid 0x10<host>
loop txqueuelen 1000 (Local Loopback)
RX packets 1978 bytes 538618 (538.6 KB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 1978 bytes 538618 (538.6 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet6 fe80::884d:baff:fe98:9c38 prefixlen 64 scopeid 0x20<link>
ether 8a:4d:ba:98:9c:38 txqueuelen 1000 (Ethernet)
RX packets 5181 bytes 1525244 (1.5 MB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 74004 bytes 7346087 (7.3 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap1: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 6a:29:52:18:5a:14 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap2: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:90:0b:15:c2:7c txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap3: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 86:ae:14:ac:75:27 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap4: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether e6:d7:b6:a4:6e:2b txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap5: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 12:2f:e4:99:a2:70 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap6: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:21:92:ad:88:a0 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap7: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 9e:fd:b7:cd:a3:e6 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
virbr0: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
inet 192.168.122.1 netmask 255.255.255.0 broadcast 192.168.122.255
ether 52:54:00:76:61:73 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
ip addr:
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default
qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: enp4s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether aa:00:04:00:9a:7c brd ff:ff:ff:ff:ff:ff permaddr 00:c0:08:97:1a:2a
3: wlp3s0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen
1000
link/ether 54:14:f3:52:56:9d brd ff:ff:ff:ff:ff:ff
4: br0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group
default qlen 1000
link/ether 3a:22:1a:4c:ea:5a brd ff:ff:ff:ff:ff:ff
inet 192.168.1.228/24 brd 192.168.1.255 scope global dynamic noprefixroute br0
valid_lft 51908sec preferred_lft 51908sec
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c/64 scope global temporary dynamic
valid_lft 7193sec preferred_lft 7193sec
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39/64 scope global dynamic mngtmpaddr
noprefixroute
valid_lft 7193sec preferred_lft 7193sec
inet6 fe80::c288:756:b4bc:213e/64 scope link noprefixroute
valid_lft forever preferred_lft forever
5: virbr0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN
group default qlen 1000
link/ether 52:54:00:76:61:73 brd ff:ff:ff:ff:ff:ff
inet 192.168.122.1/24 brd 192.168.122.255 scope global virbr0
valid_lft forever preferred_lft forever
6: tap0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether 8a:4d:ba:98:9c:38 brd ff:ff:ff:ff:ff:ff
inet6 fe80::884d:baff:fe98:9c38/64 scope link
valid_lft forever preferred_lft forever
7: tap1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 6a:29:52:18:5a:14 brd ff:ff:ff:ff:ff:ff
8: tap2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:90:0b:15:c2:7c brd ff:ff:ff:ff:ff:ff
9: tap3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 86:ae:14:ac:75:27 brd ff:ff:ff:ff:ff:ff
10: tap4: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether e6:d7:b6:a4:6e:2b brd ff:ff:ff:ff:ff:ff
11: tap5: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 12:2f:e4:99:a2:70 brd ff:ff:ff:ff:ff:ff
12: tap6: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:21:92:ad:88:a0 brd ff:ff:ff:ff:ff:ff
13: tap7: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 9e:fd:b7:cd:a3:e6 brd ff:ff:ff:ff:ff:ff
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 5:05 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
What does your physical topology look like? You have TSRTR with routing between an
incoming Multinet connection and tap0, which is a
software construct so your VAX emulators can connect to HECnet. What is TSUX1A and how is
it connected to TSRTR? The output of “ifconfig”
or “ip addr” would probably be useful here - executed on both TSRTR and TSUX1A.
John.
On Oct 25, 2023, at 3:41 PM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
Here is my tsrtr.conf for pydecnet
circuit mul-0 Multinet <hecnet gw address>:connect
circuit tap-0 Ethernet tap:tap0
routing 31.153 --type l1router --bct1=20
nsp --qmax=5
node 31.153 TSRTR
node @nodenames.dat
system --ident "TSRTR PyDECnet router (Ubuntu system)"
I did not enable web interface into TSRTR server. I need to add circuit to connect an
end-node decnet for TSUX1A.
I have br0 and enp4s0 interface. Which interface should be added so that TSUX1A can see
TSRTR router?
I have tap-0 line so that VAX emulators can see HECnet world.
Tim
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 3:08 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
On Oct 25, 2023, at 1:43 PM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
John,
Ok. I first thought MIM:: is 1.1::. I now mean 1.13::
systemctl status decnet3
● decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: active (exited) since Wed 2023-10-25 13:24:51 EDT; 1min 2s ago
Process: 1089 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1122 ExecStart=/usr/local/sbin/dnetd (code=exited, status=0/SUCCES>
Main PID: 1122 (code=exited, status=0/SUCCESS)
Tasks: 1 (limit: 37150)
Memory: 4.1M
CPU: 47ms
CGroup: /system.slice/decnet3.service
└─1133 /usr/local/sbin/dnetd
Oct 25 13:24:51 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 13:24:51 tsux1a systemd[1]: Finished decnet3.service - Load DECnet modul>
That looks good.
dneigh
Node HWtype HWaddress Flags MTU Iface
TSUX1A loop AA:00:04:00:9A:7C --- 65533 lo
That also looks good, although there should be a second line with the designated router.
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line enp4s0
OK.
I tried to add tsrtr to decnet.conf but dneigh showed that it routes to enp4s0 instead of
br0.
This release is an end-node only implementation. There can only be a single “executor”
line which defines the local name, address and
physical (ethernet or wifi) device. All other lines define node address/name pairs. There
is no concept of setting a route to a node like in TCP/IP.
TSUX1A is expected to learn about TSRTR via a multicast message on the LAN connecting
TSRTR and TSUX1A.
I tried to configure br0 instead of enp4s0 but it crashed during booting. systemctl
showed that loading failure with attempting connect to br0 interface.
Yes, that seems reasonable.
It still can't see my pydecnet router (tsrtr 31.153).
Do you know how to configure decnet with br0 interface? How to set up a route to TSRTR
(31.153)?
I think the main problem here is that TSUX1A is not seeing TSRTR. How are they connected?
What does the WEB interface into TSRTR show?
John.
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 12:17 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
On Oct 25, 2023, at 10:47 AM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
I successfully installed Linux DECnet (beta test) on Ubuntu 23.10 but have some problems
with that.
I executed 'sudo ./BuildAndInstall.sh' and followed instructions to enter, etc.
Also I started pydecnet (route to HECnet network) as bridge.
I tried to set host to my VAX simulator but got no route to host errors.
Also I tried to access 1.1:: but got no route host errors as well. Look that below.
dndir 1.1::
connect failed: No route to host
I see this as well.
sethost tsvmsb
sethost V1.0.4
Connecting to tsvmsb
connect (cterm): No route to host
ct_setup_link: error 0 connecting to host
I don’t have a VMS system set up to try this but I get further trying to connect to
RSX-11m+:
forecast@emulat:~$ sethost 41.200
sethost V1.0.4
Connecting to 41.200
ct_setup_link: error sending init sequence
sudo ncp show known nodes
Version mismatch
You don’t need the “sudo” for normal operation. The “Version mismatch” is in response to
the connection completing with the wrong
optional data. Did this system have the previous DECnet for Linux installed? If so, did
you follow the section in README.DECnet
about deleting old libraries.
sudo ncp copy known nodes from 1.1
Unable to connect to listener
Same problem as dndir above.
lsmod | grep decnet3
decnet3 86016 5
That looks good. What about the output from “systemctl status decnet3”
John.
Do you have any solutions with those problems? On my VAX simulator (TSVMSB), I was able
access HECnet without any problems.
I tried to set host to Ubuntu 23.10 but immediately got error - network is unreachable.
Tim
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Tuesday, October 24, 2023 10:50 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to
rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was
alway marked as
“experimental” and I never tried to get it running. If you need a DECnet
router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux
versions
were changes to the kernel framework APIs. By limiting the use of these
APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel
changes. So far,
I have had to make one additional conditional code change at kernel 6.5
and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal” is
still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management listener.
- There is now a subset implementation of “ncp” which is sufficient to support all
of the requests
available from the Linux network management listener. It does implement a “tell”
prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node
name database.
- The new kernel module now supports node counters as defined in the Network
Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi
releases. The current
procedure supports installation on systems derived from Debian or Fedora
although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
1:35 a.m.
What does /etc/decnet.conf look like? In order to rebuild the configuration files you need
to answer “1” to the following question:
There appears to be an existing DECnet configuration present
Do you want to:
1 - Delete the existing configuration files and create new ones
2 - Use the existing configuration files
Enter code (1 - 2):
otherwise it will reuse the old configuration files.
John.
On Oct 25, 2023, at 6:59 PM, Timothy Stark
<fsword007(a)gmail.com> wrote:
No, it did not work. TSRTR and VAX emulators can't see HECnet world with enp4s0. I
tried with br0 and it worked fine. I do not need tap0 anymore.
I tried to change to br0 by re-build configuration files with Linux DECnet. decnet3
crashed when it attempted to connect br0 interface. It said node_address is missing.
I created br0 with using brctl commands. When I upgraded Ubuntu systems for some years,
it preserved br0 configurations.
Tim
I changed a line in tsrtr.conf from pcap:tap0 line:
circuit eth-0 Ethernet pcap:br0
$ systemctl status decnet3
× decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: failed (Result: exit-code) since Wed 2023-10-25 18:29:05 EDT; 58s >
Process: 1097 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1140 ExecStart=/usr/local/sbin/dnetd (code=exited, status=255/EXCE>
Main PID: 1140 (code=exited, status=255/EXCEPTION)
CPU: 43ms
Oct 25 18:29:05 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 18:29:05 tsux1a dnetd[1140]: /proc/sys/net/decnet/node_address missing, >
Oct 25 18:29:05 tsux1a systemd[1]: decnet3.service: Main process exited, code=e>
Oct 25 18:29:05 tsux1a systemd[1]: decnet3.service: Failed with result 'exit-co>
Oct 25 18:29:05 tsux1a systemd[1]: Failed to start decnet3.service - Load DECne>
$ dneigh
Error: can not open DECnet neigh file: /proc/net/decnet_neigh: No such file or directory
$ brctl show
bridge name bridge id STP enabled interfaces
br0 8000.3a221a4cea5a yes enp4s0
virbr0 8000.525400766173 yes
$ cat /proc/net/decnet_dev
Name Flags T1 Timer1 T3 Timer3 T4 Timer4 BlkSize Pri State DevType Router
Peer
lo B 0000 0000 0000 0000 0000 0000 65533 000 00 loopback
br0 B 0000 0000 0000 0010 0000 0000 1498 000 00 ethernet
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 5:53 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
I’m not familiar with the FIOS router configuration. I’m going to guess that there’s
one or more ethernet’s coming out of it; one going to TSRTR
and another going to TSUX1A. On TSRTR that is probably “enp4s0”. You will need to add a
circuit to pydecnet so it does routing on your
ethernet segment, something like:
circuit eth-0 Ethernet pcap:enp4s0
John.
On Oct 25, 2023, at 5:29 PM, Timothy Stark
<fsword007(a)gmail.com <mailto:fsword007@gmail.com>> wrote:
John,
I set virtual br0 for 8 tap devices that VAX emulators and others use. br0 is the main
routing. virbr0 is for QEMU VM (OpenVMS x86). TSUX1A is IP 192.168.1.228.
TSUX1A is Ubuntu 23.10 system on my embedded box. TSRTR is pydecnet router for HECnet
world access that connected to tap0. I am figuring out why TSRTR and TSUX1A can't see
each other. I am only able access TSVMSB (on SIMH VAX emulator) that uses tap1 through
TCP/IP but not DECnet.
TSUX1A is connected to my Verizon FiOS router for Internet access.
Here is ifconfig output below.
Tim
br0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 192.168.1.228 netmask 255.255.255.0 broadcast 192.168.1.255
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c prefixlen 64 scopeid
0x0<global>
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39 prefixlen 64 scopeid
0x0<global>
inet6 fe80::c288:756:b4bc:213e prefixlen 64 scopeid 0x20<link>
ether 3a:22:1a:4c:ea:5a txqueuelen 1000 (Ethernet)
RX packets 85254 bytes 10370418 (10.3 MB)
RX errors 0 dropped 8364 overruns 0 frame 0
TX packets 8096 bytes 813053 (813.0 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
enp4s0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
ether aa:00:04:00:9a:7c txqueuelen 1000 (Ethernet)
RX packets 100400 bytes 16878747 (16.8 MB)
RX errors 0 dropped 14 overruns 0 frame 0
TX packets 21289 bytes 2748243 (2.7 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
lo: flags=73<UP,LOOPBACK,RUNNING> mtu 65536
inet 127.0.0.1 netmask 255.0.0.0
inet6 ::1 prefixlen 128 scopeid 0x10<host>
loop txqueuelen 1000 (Local Loopback)
RX packets 1978 bytes 538618 (538.6 KB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 1978 bytes 538618 (538.6 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet6 fe80::884d:baff:fe98:9c38 prefixlen 64 scopeid 0x20<link>
ether 8a:4d:ba:98:9c:38 txqueuelen 1000 (Ethernet)
RX packets 5181 bytes 1525244 (1.5 MB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 74004 bytes 7346087 (7.3 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap1: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 6a:29:52:18:5a:14 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap2: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:90:0b:15:c2:7c txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap3: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 86:ae:14:ac:75:27 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap4: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether e6:d7:b6:a4:6e:2b txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap5: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 12:2f:e4:99:a2:70 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap6: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:21:92:ad:88:a0 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap7: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 9e:fd:b7:cd:a3:e6 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
virbr0: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
inet 192.168.122.1 netmask 255.255.255.0 broadcast 192.168.122.255
ether 52:54:00:76:61:73 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
ip addr:
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default
qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: enp4s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether aa:00:04:00:9a:7c brd ff:ff:ff:ff:ff:ff permaddr 00:c0:08:97:1a:2a
3: wlp3s0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen
1000
link/ether 54:14:f3:52:56:9d brd ff:ff:ff:ff:ff:ff
4: br0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group
default qlen 1000
link/ether 3a:22:1a:4c:ea:5a brd ff:ff:ff:ff:ff:ff
inet 192.168.1.228/24 brd 192.168.1.255 scope global dynamic noprefixroute br0
valid_lft 51908sec preferred_lft 51908sec
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c/64 scope global temporary dynamic
valid_lft 7193sec preferred_lft 7193sec
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39/64 scope global dynamic mngtmpaddr
noprefixroute
valid_lft 7193sec preferred_lft 7193sec
inet6 fe80::c288:756:b4bc:213e/64 scope link noprefixroute
valid_lft forever preferred_lft forever
5: virbr0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN
group default qlen 1000
link/ether 52:54:00:76:61:73 brd ff:ff:ff:ff:ff:ff
inet 192.168.122.1/24 brd 192.168.122.255 scope global virbr0
valid_lft forever preferred_lft forever
6: tap0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether 8a:4d:ba:98:9c:38 brd ff:ff:ff:ff:ff:ff
inet6 fe80::884d:baff:fe98:9c38/64 scope link
valid_lft forever preferred_lft forever
7: tap1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 6a:29:52:18:5a:14 brd ff:ff:ff:ff:ff:ff
8: tap2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:90:0b:15:c2:7c brd ff:ff:ff:ff:ff:ff
9: tap3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 86:ae:14:ac:75:27 brd ff:ff:ff:ff:ff:ff
10: tap4: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether e6:d7:b6:a4:6e:2b brd ff:ff:ff:ff:ff:ff
11: tap5: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 12:2f:e4:99:a2:70 brd ff:ff:ff:ff:ff:ff
12: tap6: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:21:92:ad:88:a0 brd ff:ff:ff:ff:ff:ff
13: tap7: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 9e:fd:b7:cd:a3:e6 brd ff:ff:ff:ff:ff:ff
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 5:05 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
What does your physical topology look like? You have TSRTR with routing between an
incoming Multinet connection and tap0, which is a
software construct so your VAX emulators can connect to HECnet. What is TSUX1A and how is
it connected to TSRTR? The output of “ifconfig”
or “ip addr” would probably be useful here - executed on both TSRTR and TSUX1A.
John.
On Oct 25, 2023, at 3:41 PM, Timothy Stark <fsword007(a)gmail.com
<mailto:fsword007@gmail.com>> wrote:
Here is my tsrtr.conf for pydecnet
circuit mul-0 Multinet <hecnet gw address>:connect
circuit tap-0 Ethernet tap:tap0
routing 31.153 --type l1router --bct1=20
nsp --qmax=5
node 31.153 TSRTR
node @nodenames.dat
system --ident "TSRTR PyDECnet router (Ubuntu system)"
I did not enable web interface into TSRTR server. I need to add circuit to connect an
end-node decnet for TSUX1A.
I have br0 and enp4s0 interface. Which interface should be added so that TSUX1A can see
TSRTR router?
I have tap-0 line so that VAX emulators can see HECnet world.
Tim
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 3:08 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
On Oct 25, 2023, at 1:43 PM, Timothy Stark <fsword007(a)gmail.com
<mailto:fsword007@gmail.com>> wrote:
John,
Ok. I first thought MIM:: is 1.1::. I now mean 1.13::
systemctl status decnet3
● decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: active (exited) since Wed 2023-10-25 13:24:51 EDT; 1min 2s ago
Process: 1089 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1122 ExecStart=/usr/local/sbin/dnetd (code=exited, status=0/SUCCES>
Main PID: 1122 (code=exited, status=0/SUCCESS)
Tasks: 1 (limit: 37150)
Memory: 4.1M
CPU: 47ms
CGroup: /system.slice/decnet3.service
└─1133 /usr/local/sbin/dnetd
Oct 25 13:24:51 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 13:24:51 tsux1a systemd[1]: Finished decnet3.service - Load DECnet modul>
That looks good.
dneigh
Node HWtype HWaddress Flags MTU Iface
TSUX1A loop AA:00:04:00:9A:7C --- 65533 lo
That also looks good, although there should be a second line with the designated router.
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line enp4s0
OK.
I tried to add tsrtr to decnet.conf but dneigh showed that it routes to enp4s0 instead of
br0.
This release is an end-node only implementation. There can only be a single “executor”
line which defines the local name, address and
physical (ethernet or wifi) device. All other lines define node address/name pairs. There
is no concept of setting a route to a node like in TCP/IP.
TSUX1A is expected to learn about TSRTR via a multicast message on the LAN connecting
TSRTR and TSUX1A.
I tried to configure br0 instead of enp4s0 but it crashed during booting. systemctl
showed that loading failure with attempting connect to br0 interface.
Yes, that seems reasonable.
It still can't see my pydecnet router (tsrtr 31.153).
Do you know how to configure decnet with br0 interface? How to set up a route to TSRTR
(31.153)?
I think the main problem here is that TSUX1A is not seeing TSRTR. How are they connected?
What does the WEB interface into TSRTR show?
John.
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 12:17 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
On Oct 25, 2023, at 10:47 AM, Timothy Stark <fsword007(a)gmail.com
<mailto:fsword007@gmail.com>> wrote:
I successfully installed Linux DECnet (beta test) on Ubuntu 23.10 but have some problems
with that.
I executed 'sudo ./BuildAndInstall.sh' and followed instructions to enter, etc.
Also I started pydecnet (route to HECnet network) as bridge.
I tried to set host to my VAX simulator but got no route to host errors.
Also I tried to access 1.1:: but got no route host errors as well. Look that below.
dndir 1.1::
connect failed: No route to host
I see this as well.
sethost tsvmsb
sethost V1.0.4
Connecting to tsvmsb
connect (cterm): No route to host
ct_setup_link: error 0 connecting to host
I don’t have a VMS system set up to try this but I get further trying to connect to
RSX-11m+:
forecast@emulat:~$ sethost 41.200
sethost V1.0.4
Connecting to 41.200
ct_setup_link: error sending init sequence
sudo ncp show known nodes
Version mismatch
You don’t need the “sudo” for normal operation. The “Version mismatch” is in response to
the connection completing with the wrong
optional data. Did this system have the previous DECnet for Linux installed? If so, did
you follow the section in README.DECnet
about deleting old libraries.
sudo ncp copy known nodes from 1.1
Unable to connect to listener
Same problem as dndir above.
lsmod | grep decnet3
decnet3 86016 5
That looks good. What about the output from “systemctl status decnet3”
John.
Do you have any solutions with those problems? On my VAX simulator (TSVMSB), I was able
access HECnet without any problems.
I tried to set host to Ubuntu 23.10 but immediately got error - network is unreachable.
Tim
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Tuesday, October 24, 2023 10:50 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet
<https://github.com/JohnForecast/RaspbianDECnet>> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to
rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was
alway marked as
“experimental” and I never tried to get it running. If you need a DECnet
router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux
versions
were changes to the kernel framework APIs. By limiting the use of these
APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel
changes. So far,
I have had to make one additional conditional code change at kernel 6.5
and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal”
is still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management
listener.
- There is now a subset implementation of “ncp” which is sufficient to support
all of the requests
available from the Linux network management listener. It does implement a
“tell” prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node
name database.
- The new kernel module now supports node counters as defined in the Network
Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi
releases. The current
procedure supports installation on systems derived from Debian or Fedora
although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet
<https://github.com/JohnForecast/LinuxDECnet>>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
1:53 a.m.
John,
I already did that. It still did not work.
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line br0
node 31.153 name TSRTR
Tim
________________________________
From: John Forecast <john(a)forecast.name>
Sent: Wednesday, October 25, 2023 7:35 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
What does /etc/decnet.conf look like? In order to rebuild the configuration files you need
to answer “1” to the following question:
There appears to be an existing DECnet configuration present
Do you want to:
1 - Delete the existing configuration files and create new ones
2 - Use the existing configuration files
Enter code (1 - 2):
otherwise it will reuse the old configuration files.
John.
On Oct 25, 2023, at 6:59 PM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
No, it did not work. TSRTR and VAX emulators can't see HECnet world with enp4s0. I
tried with br0 and it worked fine. I do not need tap0 anymore.
I tried to change to br0 by re-build configuration files with Linux DECnet. decnet3
crashed when it attempted to connect br0 interface. It said node_address is missing.
I created br0 with using brctl commands. When I upgraded Ubuntu systems for some years,
it preserved br0 configurations.
Tim
I changed a line in tsrtr.conf from pcap:tap0 line:
circuit eth-0 Ethernet pcap:br0
$ systemctl status decnet3
× decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: failed (Result: exit-code) since Wed 2023-10-25 18:29:05 EDT; 58s >
Process: 1097 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1140 ExecStart=/usr/local/sbin/dnetd (code=exited, status=255/EXCE>
Main PID: 1140 (code=exited, status=255/EXCEPTION)
CPU: 43ms
Oct 25 18:29:05 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 18:29:05 tsux1a dnetd[1140]: /proc/sys/net/decnet/node_address missing, >
Oct 25 18:29:05 tsux1a systemd[1]: decnet3.service: Main process exited, code=e>
Oct 25 18:29:05 tsux1a systemd[1]: decnet3.service: Failed with result 'exit-co>
Oct 25 18:29:05 tsux1a systemd[1]: Failed to start decnet3.service - Load DECne>
$ dneigh
Error: can not open DECnet neigh file: /proc/net/decnet_neigh: No such file or directory
$ brctl show
bridge name bridge id STP enabled interfaces
br0 8000.3a221a4cea5a yes enp4s0
virbr0 8000.525400766173 yes
$ cat /proc/net/decnet_dev
Name Flags T1 Timer1 T3 Timer3 T4 Timer4 BlkSize Pri State DevType Router
Peer
lo B 0000 0000 0000 0000 0000 0000 65533 000 00 loopback
br0 B 0000 0000 0000 0010 0000 0000 1498 000 00 ethernet
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 5:53 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
I’m not familiar with the FIOS router configuration. I’m going to guess that there’s
one or more ethernet’s coming out of it; one going to TSRTR
and another going to TSUX1A. On TSRTR that is probably “enp4s0”. You will need to add a
circuit to pydecnet so it does routing on your
ethernet segment, something like:
circuit eth-0 Ethernet pcap:enp4s0
John.
On Oct 25, 2023, at 5:29 PM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
John,
I set virtual br0 for 8 tap devices that VAX emulators and others use. br0 is the main
routing. virbr0 is for QEMU VM (OpenVMS x86). TSUX1A is IP 192.168.1.228.
TSUX1A is Ubuntu 23.10 system on my embedded box. TSRTR is pydecnet router for HECnet
world access that connected to tap0. I am figuring out why TSRTR and TSUX1A can't see
each other. I am only able access TSVMSB (on SIMH VAX emulator) that uses tap1 through
TCP/IP but not DECnet.
TSUX1A is connected to my Verizon FiOS router for Internet access.
Here is ifconfig output below.
Tim
br0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 192.168.1.228 netmask 255.255.255.0 broadcast 192.168.1.255
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c prefixlen 64 scopeid
0x0<global>
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39 prefixlen 64 scopeid
0x0<global>
inet6 fe80::c288:756:b4bc:213e prefixlen 64 scopeid 0x20<link>
ether 3a:22:1a:4c:ea:5a txqueuelen 1000 (Ethernet)
RX packets 85254 bytes 10370418 (10.3 MB)
RX errors 0 dropped 8364 overruns 0 frame 0
TX packets 8096 bytes 813053 (813.0 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
enp4s0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
ether aa:00:04:00:9a:7c txqueuelen 1000 (Ethernet)
RX packets 100400 bytes 16878747 (16.8 MB)
RX errors 0 dropped 14 overruns 0 frame 0
TX packets 21289 bytes 2748243 (2.7 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
lo: flags=73<UP,LOOPBACK,RUNNING> mtu 65536
inet 127.0.0.1 netmask 255.0.0.0
inet6 ::1 prefixlen 128 scopeid 0x10<host>
loop txqueuelen 1000 (Local Loopback)
RX packets 1978 bytes 538618 (538.6 KB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 1978 bytes 538618 (538.6 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet6 fe80::884d:baff:fe98:9c38 prefixlen 64 scopeid 0x20<link>
ether 8a:4d:ba:98:9c:38 txqueuelen 1000 (Ethernet)
RX packets 5181 bytes 1525244 (1.5 MB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 74004 bytes 7346087 (7.3 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap1: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 6a:29:52:18:5a:14 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap2: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:90:0b:15:c2:7c txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap3: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 86:ae:14:ac:75:27 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap4: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether e6:d7:b6:a4:6e:2b txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap5: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 12:2f:e4:99:a2:70 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap6: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:21:92:ad:88:a0 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap7: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 9e:fd:b7:cd:a3:e6 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
virbr0: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
inet 192.168.122.1 netmask 255.255.255.0 broadcast 192.168.122.255
ether 52:54:00:76:61:73 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
ip addr:
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default
qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: enp4s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether aa:00:04:00:9a:7c brd ff:ff:ff:ff:ff:ff permaddr 00:c0:08:97:1a:2a
3: wlp3s0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen
1000
link/ether 54:14:f3:52:56:9d brd ff:ff:ff:ff:ff:ff
4: br0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group
default qlen 1000
link/ether 3a:22:1a:4c:ea:5a brd ff:ff:ff:ff:ff:ff
inet 192.168.1.228/24 brd 192.168.1.255 scope global dynamic noprefixroute br0
valid_lft 51908sec preferred_lft 51908sec
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c/64 scope global temporary dynamic
valid_lft 7193sec preferred_lft 7193sec
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39/64 scope global dynamic mngtmpaddr
noprefixroute
valid_lft 7193sec preferred_lft 7193sec
inet6 fe80::c288:756:b4bc:213e/64 scope link noprefixroute
valid_lft forever preferred_lft forever
5: virbr0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN
group default qlen 1000
link/ether 52:54:00:76:61:73 brd ff:ff:ff:ff:ff:ff
inet 192.168.122.1/24 brd 192.168.122.255 scope global virbr0
valid_lft forever preferred_lft forever
6: tap0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether 8a:4d:ba:98:9c:38 brd ff:ff:ff:ff:ff:ff
inet6 fe80::884d:baff:fe98:9c38/64 scope link
valid_lft forever preferred_lft forever
7: tap1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 6a:29:52:18:5a:14 brd ff:ff:ff:ff:ff:ff
8: tap2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:90:0b:15:c2:7c brd ff:ff:ff:ff:ff:ff
9: tap3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 86:ae:14:ac:75:27 brd ff:ff:ff:ff:ff:ff
10: tap4: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether e6:d7:b6:a4:6e:2b brd ff:ff:ff:ff:ff:ff
11: tap5: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 12:2f:e4:99:a2:70 brd ff:ff:ff:ff:ff:ff
12: tap6: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:21:92:ad:88:a0 brd ff:ff:ff:ff:ff:ff
13: tap7: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 9e:fd:b7:cd:a3:e6 brd ff:ff:ff:ff:ff:ff
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 5:05 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
What does your physical topology look like? You have TSRTR with routing between an
incoming Multinet connection and tap0, which is a
software construct so your VAX emulators can connect to HECnet. What is TSUX1A and how is
it connected to TSRTR? The output of “ifconfig”
or “ip addr” would probably be useful here - executed on both TSRTR and TSUX1A.
John.
On Oct 25, 2023, at 3:41 PM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
Here is my tsrtr.conf for pydecnet
circuit mul-0 Multinet <hecnet gw address>:connect
circuit tap-0 Ethernet tap:tap0
routing 31.153 --type l1router --bct1=20
nsp --qmax=5
node 31.153 TSRTR
node @nodenames.dat
system --ident "TSRTR PyDECnet router (Ubuntu system)"
I did not enable web interface into TSRTR server. I need to add circuit to connect an
end-node decnet for TSUX1A.
I have br0 and enp4s0 interface. Which interface should be added so that TSUX1A can see
TSRTR router?
I have tap-0 line so that VAX emulators can see HECnet world.
Tim
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 3:08 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
On Oct 25, 2023, at 1:43 PM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
John,
Ok. I first thought MIM:: is 1.1::. I now mean 1.13::
systemctl status decnet3
● decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: active (exited) since Wed 2023-10-25 13:24:51 EDT; 1min 2s ago
Process: 1089 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1122 ExecStart=/usr/local/sbin/dnetd (code=exited, status=0/SUCCES>
Main PID: 1122 (code=exited, status=0/SUCCESS)
Tasks: 1 (limit: 37150)
Memory: 4.1M
CPU: 47ms
CGroup: /system.slice/decnet3.service
└─1133 /usr/local/sbin/dnetd
Oct 25 13:24:51 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 13:24:51 tsux1a systemd[1]: Finished decnet3.service - Load DECnet modul>
That looks good.
dneigh
Node HWtype HWaddress Flags MTU Iface
TSUX1A loop AA:00:04:00:9A:7C --- 65533 lo
That also looks good, although there should be a second line with the designated router.
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line enp4s0
OK.
I tried to add tsrtr to decnet.conf but dneigh showed that it routes to enp4s0 instead of
br0.
This release is an end-node only implementation. There can only be a single “executor”
line which defines the local name, address and
physical (ethernet or wifi) device. All other lines define node address/name pairs. There
is no concept of setting a route to a node like in TCP/IP.
TSUX1A is expected to learn about TSRTR via a multicast message on the LAN connecting
TSRTR and TSUX1A.
I tried to configure br0 instead of enp4s0 but it crashed during booting. systemctl
showed that loading failure with attempting connect to br0 interface.
Yes, that seems reasonable.
It still can't see my pydecnet router (tsrtr 31.153).
Do you know how to configure decnet with br0 interface? How to set up a route to TSRTR
(31.153)?
I think the main problem here is that TSUX1A is not seeing TSRTR. How are they connected?
What does the WEB interface into TSRTR show?
John.
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 12:17 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
On Oct 25, 2023, at 10:47 AM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
I successfully installed Linux DECnet (beta test) on Ubuntu 23.10 but have some problems
with that.
I executed 'sudo ./BuildAndInstall.sh' and followed instructions to enter, etc.
Also I started pydecnet (route to HECnet network) as bridge.
I tried to set host to my VAX simulator but got no route to host errors.
Also I tried to access 1.1:: but got no route host errors as well. Look that below.
dndir 1.1::
connect failed: No route to host
I see this as well.
sethost tsvmsb
sethost V1.0.4
Connecting to tsvmsb
connect (cterm): No route to host
ct_setup_link: error 0 connecting to host
I don’t have a VMS system set up to try this but I get further trying to connect to
RSX-11m+:
forecast@emulat:~$ sethost 41.200
sethost V1.0.4
Connecting to 41.200
ct_setup_link: error sending init sequence
sudo ncp show known nodes
Version mismatch
You don’t need the “sudo” for normal operation. The “Version mismatch” is in response to
the connection completing with the wrong
optional data. Did this system have the previous DECnet for Linux installed? If so, did
you follow the section in README.DECnet
about deleting old libraries.
sudo ncp copy known nodes from 1.1
Unable to connect to listener
Same problem as dndir above.
lsmod | grep decnet3
decnet3 86016 5
That looks good. What about the output from “systemctl status decnet3”
John.
Do you have any solutions with those problems? On my VAX simulator (TSVMSB), I was able
access HECnet without any problems.
I tried to set host to Ubuntu 23.10 but immediately got error - network is unreachable.
Tim
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Tuesday, October 24, 2023 10:50 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to
rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was
alway marked as
“experimental” and I never tried to get it running. If you need a DECnet
router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux
versions
were changes to the kernel framework APIs. By limiting the use of these
APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel
changes. So far,
I have had to make one additional conditional code change at kernel 6.5
and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal” is
still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management listener.
- There is now a subset implementation of “ncp” which is sufficient to support all
of the requests
available from the Linux network management listener. It does implement a “tell”
prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node
name database.
- The new kernel module now supports node counters as defined in the Network
Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi
releases. The current
procedure supports installation on systems derived from Debian or Fedora
although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
2:30 a.m.
Tim,
Bridges seem to interact badly with systemd. systemd does not wait for the bridge to
come up, it only waits for the physical ethernet device.
The result is that when DECnet tries to change the MAC address of br0 it does not exist.
The simplest way to make this work is to set the bridge’s MAC address when you create it.
Here’s my configuration file for creating br0:
auto br0
iface br0 inet static
bridge_hw aa:00:04:00:ca:a4
address 192.168.1.39/24
gateway 192.168.1.1
bridge_ports eth0 tap0
bridge_stp off
bridge_maxwait 5
dns-nameservers 192.168.1.12
Yours should read:
bridge_hw aa:00:04:00:9a:7c
If you use explicit brctl commands to set up your bridge, you’ll have to add the
following command after initially creating br0:
ip link set br0 address aa:00:04:00:9a:7c
John.
On Oct 25, 2023, at 7:53 PM, Timothy Stark
<fsword007(a)gmail.com> wrote:
John,
I already did that. It still did not work.
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line br0
node 31.153 name TSRTR
Tim
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 7:35 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
What does /etc/decnet.conf look like? In order to rebuild the configuration files you
need to answer “1” to the following question:
There appears to be an existing DECnet configuration present
Do you want to:
1 - Delete the existing configuration files and create new ones
2 - Use the existing configuration files
Enter code (1 - 2):
otherwise it will reuse the old configuration files.
John.
On Oct 25, 2023, at 6:59 PM, Timothy Stark
<fsword007(a)gmail.com <mailto:fsword007@gmail.com>> wrote:
No, it did not work. TSRTR and VAX emulators can't see HECnet world with enp4s0. I
tried with br0 and it worked fine. I do not need tap0 anymore.
I tried to change to br0 by re-build configuration files with Linux DECnet. decnet3
crashed when it attempted to connect br0 interface. It said node_address is missing.
I created br0 with using brctl commands. When I upgraded Ubuntu systems for some years,
it preserved br0 configurations.
Tim
I changed a line in tsrtr.conf from pcap:tap0 line:
circuit eth-0 Ethernet pcap:br0
$ systemctl status decnet3
× decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: failed (Result: exit-code) since Wed 2023-10-25 18:29:05 EDT; 58s >
Process: 1097 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1140 ExecStart=/usr/local/sbin/dnetd (code=exited, status=255/EXCE>
Main PID: 1140 (code=exited, status=255/EXCEPTION)
CPU: 43ms
Oct 25 18:29:05 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 18:29:05 tsux1a dnetd[1140]: /proc/sys/net/decnet/node_address missing, >
Oct 25 18:29:05 tsux1a systemd[1]: decnet3.service: Main process exited, code=e>
Oct 25 18:29:05 tsux1a systemd[1]: decnet3.service: Failed with result 'exit-co>
Oct 25 18:29:05 tsux1a systemd[1]: Failed to start decnet3.service - Load DECne>
$ dneigh
Error: can not open DECnet neigh file: /proc/net/decnet_neigh: No such file or directory
$ brctl show
bridge name bridge id STP enabled interfaces
br0 8000.3a221a4cea5a yes enp4s0
virbr0 8000.525400766173 yes
$ cat /proc/net/decnet_dev
Name Flags T1 Timer1 T3 Timer3 T4 Timer4 BlkSize Pri State DevType Router
Peer
lo B 0000 0000 0000 0000 0000 0000 65533 000 00 loopback
br0 B 0000 0000 0000 0010 0000 0000 1498 000 00 ethernet
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 5:53 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
I’m not familiar with the FIOS router configuration. I’m going to guess that there’s
one or more ethernet’s coming out of it; one going to TSRTR
and another going to TSUX1A. On TSRTR that is probably “enp4s0”. You will need to add a
circuit to pydecnet so it does routing on your
ethernet segment, something like:
circuit eth-0 Ethernet pcap:enp4s0
John.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se> On Oct 25, 2023, at 5:29 PM, Timothy Stark
<fsword007(a)gmail.com <mailto:fsword007@gmail.com>> wrote:
John,
I set virtual br0 for 8 tap devices that VAX emulators and others use. br0 is the main
routing. virbr0 is for QEMU VM (OpenVMS x86). TSUX1A is IP 192.168.1.228.
TSUX1A is Ubuntu 23.10 system on my embedded box. TSRTR is pydecnet router for HECnet
world access that connected to tap0. I am figuring out why TSRTR and TSUX1A can't see
each other. I am only able access TSVMSB (on SIMH VAX emulator) that uses tap1 through
TCP/IP but not DECnet.
TSUX1A is connected to my Verizon FiOS router for Internet access.
Here is ifconfig output below.
Tim
br0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 192.168.1.228 netmask 255.255.255.0 broadcast 192.168.1.255
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c prefixlen 64 scopeid
0x0<global>
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39 prefixlen 64 scopeid
0x0<global>
inet6 fe80::c288:756:b4bc:213e prefixlen 64 scopeid 0x20<link>
ether 3a:22:1a:4c:ea:5a txqueuelen 1000 (Ethernet)
RX packets 85254 bytes 10370418 (10.3 MB)
RX errors 0 dropped 8364 overruns 0 frame 0
TX packets 8096 bytes 813053 (813.0 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
enp4s0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
ether aa:00:04:00:9a:7c txqueuelen 1000 (Ethernet)
RX packets 100400 bytes 16878747 (16.8 MB)
RX errors 0 dropped 14 overruns 0 frame 0
TX packets 21289 bytes 2748243 (2.7 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
lo: flags=73<UP,LOOPBACK,RUNNING> mtu 65536
inet 127.0.0.1 netmask 255.0.0.0
inet6 ::1 prefixlen 128 scopeid 0x10<host>
loop txqueuelen 1000 (Local Loopback)
RX packets 1978 bytes 538618 (538.6 KB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 1978 bytes 538618 (538.6 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet6 fe80::884d:baff:fe98:9c38 prefixlen 64 scopeid 0x20<link>
ether 8a:4d:ba:98:9c:38 txqueuelen 1000 (Ethernet)
RX packets 5181 bytes 1525244 (1.5 MB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 74004 bytes 7346087 (7.3 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap1: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 6a:29:52:18:5a:14 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap2: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:90:0b:15:c2:7c txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap3: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 86:ae:14:ac:75:27 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap4: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether e6:d7:b6:a4:6e:2b txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap5: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 12:2f:e4:99:a2:70 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap6: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:21:92:ad:88:a0 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap7: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 9e:fd:b7:cd:a3:e6 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
virbr0: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
inet 192.168.122.1 netmask 255.255.255.0 broadcast 192.168.122.255
ether 52:54:00:76:61:73 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
ip addr:
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default
qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: enp4s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether aa:00:04:00:9a:7c brd ff:ff:ff:ff:ff:ff permaddr 00:c0:08:97:1a:2a
3: wlp3s0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen
1000
link/ether 54:14:f3:52:56:9d brd ff:ff:ff:ff:ff:ff
4: br0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group
default qlen 1000
link/ether 3a:22:1a:4c:ea:5a brd ff:ff:ff:ff:ff:ff
inet 192.168.1.228/24 brd 192.168.1.255 scope global dynamic noprefixroute br0
valid_lft 51908sec preferred_lft 51908sec
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c/64 scope global temporary dynamic
valid_lft 7193sec preferred_lft 7193sec
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39/64 scope global dynamic mngtmpaddr
noprefixroute
valid_lft 7193sec preferred_lft 7193sec
inet6 fe80::c288:756:b4bc:213e/64 scope link noprefixroute
valid_lft forever preferred_lft forever
5: virbr0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN
group default qlen 1000
link/ether 52:54:00:76:61:73 brd ff:ff:ff:ff:ff:ff
inet 192.168.122.1/24 brd 192.168.122.255 scope global virbr0
valid_lft forever preferred_lft forever
6: tap0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether 8a:4d:ba:98:9c:38 brd ff:ff:ff:ff:ff:ff
inet6 fe80::884d:baff:fe98:9c38/64 scope link
valid_lft forever preferred_lft forever
7: tap1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 6a:29:52:18:5a:14 brd ff:ff:ff:ff:ff:ff
8: tap2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:90:0b:15:c2:7c brd ff:ff:ff:ff:ff:ff
9: tap3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 86:ae:14:ac:75:27 brd ff:ff:ff:ff:ff:ff
10: tap4: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether e6:d7:b6:a4:6e:2b brd ff:ff:ff:ff:ff:ff
11: tap5: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 12:2f:e4:99:a2:70 brd ff:ff:ff:ff:ff:ff
12: tap6: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:21:92:ad:88:a0 brd ff:ff:ff:ff:ff:ff
13: tap7: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 9e:fd:b7:cd:a3:e6 brd ff:ff:ff:ff:ff:ff
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 5:05 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
What does your physical topology look like? You have TSRTR with routing between an
incoming Multinet connection and tap0, which is a
software construct so your VAX emulators can connect to HECnet. What is TSUX1A and how is
it connected to TSRTR? The output of “ifconfig”
or “ip addr” would probably be useful here - executed on both TSRTR and TSUX1A.
John.
On Oct 25, 2023, at 3:41 PM, Timothy Stark <fsword007(a)gmail.com
<mailto:fsword007@gmail.com>> wrote:
Here is my tsrtr.conf for pydecnet
circuit mul-0 Multinet <hecnet gw address>:connect
circuit tap-0 Ethernet tap:tap0
routing 31.153 --type l1router --bct1=20
nsp --qmax=5
node 31.153 TSRTR
node @nodenames.dat
system --ident "TSRTR PyDECnet router (Ubuntu system)"
I did not enable web interface into TSRTR server. I need to add circuit to connect an
end-node decnet for TSUX1A.
I have br0 and enp4s0 interface. Which interface should be added so that TSUX1A can see
TSRTR router?
I have tap-0 line so that VAX emulators can see HECnet world.
Tim
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 3:08 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
On Oct 25, 2023, at 1:43 PM, Timothy Stark <fsword007(a)gmail.com
<mailto:fsword007@gmail.com>> wrote:
John,
Ok. I first thought MIM:: is 1.1::. I now mean 1.13::
systemctl status decnet3
● decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: active (exited) since Wed 2023-10-25 13:24:51 EDT; 1min 2s ago
Process: 1089 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1122 ExecStart=/usr/local/sbin/dnetd (code=exited, status=0/SUCCES>
Main PID: 1122 (code=exited, status=0/SUCCESS)
Tasks: 1 (limit: 37150)
Memory: 4.1M
CPU: 47ms
CGroup: /system.slice/decnet3.service
└─1133 /usr/local/sbin/dnetd
Oct 25 13:24:51 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 13:24:51 tsux1a systemd[1]: Finished decnet3.service - Load DECnet modul>
That looks good.
dneigh
Node HWtype HWaddress Flags MTU Iface
TSUX1A loop AA:00:04:00:9A:7C --- 65533 lo
That also looks good, although there should be a second line with the designated router.
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line enp4s0
OK.
I tried to add tsrtr to decnet.conf but dneigh showed that it routes to enp4s0 instead of
br0.
This release is an end-node only implementation. There can only be a single “executor”
line which defines the local name, address and
physical (ethernet or wifi) device. All other lines define node address/name pairs. There
is no concept of setting a route to a node like in TCP/IP.
TSUX1A is expected to learn about TSRTR via a multicast message on the LAN connecting
TSRTR and TSUX1A.
I tried to configure br0 instead of enp4s0 but it crashed during booting. systemctl
showed that loading failure with attempting connect to br0 interface.
Yes, that seems reasonable.
It still can't see my pydecnet router (tsrtr 31.153).
Do you know how to configure decnet with br0 interface? How to set up a route to TSRTR
(31.153)?
I think the main problem here is that TSUX1A is not seeing TSRTR. How are they connected?
What does the WEB interface into TSRTR show?
John.
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 12:17 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
On Oct 25, 2023, at 10:47 AM, Timothy Stark <fsword007(a)gmail.com
<mailto:fsword007@gmail.com>> wrote:
I successfully installed Linux DECnet (beta test) on Ubuntu 23.10 but have some problems
with that.
I executed 'sudo ./BuildAndInstall.sh' and followed instructions to enter, etc.
Also I started pydecnet (route to HECnet network) as bridge.
I tried to set host to my VAX simulator but got no route to host errors.
Also I tried to access 1.1:: but got no route host errors as well. Look that below.
dndir 1.1::
connect failed: No route to host
I see this as well.
sethost tsvmsb
sethost V1.0.4
Connecting to tsvmsb
connect (cterm): No route to host
ct_setup_link: error 0 connecting to host
I don’t have a VMS system set up to try this but I get further trying to connect to
RSX-11m+:
forecast@emulat:~$ sethost 41.200
sethost V1.0.4
Connecting to 41.200
ct_setup_link: error sending init sequence
sudo ncp show known nodes
Version mismatch
You don’t need the “sudo” for normal operation. The “Version mismatch” is in response to
the connection completing with the wrong
optional data. Did this system have the previous DECnet for Linux installed? If so, did
you follow the section in README.DECnet
about deleting old libraries.
sudo ncp copy known nodes from 1.1
Unable to connect to listener
Same problem as dndir above.
lsmod | grep decnet3
decnet3 86016 5
That looks good. What about the output from “systemctl status decnet3”
John.
Do you have any solutions with those problems? On my VAX simulator (TSVMSB), I was able
access HECnet without any problems.
I tried to set host to Ubuntu 23.10 but immediately got error - network is unreachable.
Tim
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Tuesday, October 24, 2023 10:50 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet
<https://github.com/JohnForecast/RaspbianDECnet>> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to
rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was
alway marked as
“experimental” and I never tried to get it running. If you need a DECnet
router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux
versions
were changes to the kernel framework APIs. By limiting the use of these
APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel
changes. So far,
I have had to make one additional conditional code change at kernel 6.5
and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal”
is still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management
listener.
- There is now a subset implementation of “ncp” which is sufficient to support
all of the requests
available from the Linux network management listener. It does implement a
“tell” prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node
name database.
- The new kernel module now supports node counters as defined in the Network
Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi
releases. The current
procedure supports installation on systems derived from Debian or Fedora
although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet
<https://github.com/JohnForecast/LinuxDECnet>>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
1:51 p.m.
Ok, I now got it. I use NetworkManager (nmcli commands) for adding br0 because I am using
Ubuntu system. I was searching for it and found configuration files in /etc/netplan
directory. I was looking for some manual about setting hwaddr but can't find it.
I tried to ip link set br0 address but it returned back to own addr after boot. decnet3
still failed to load.
I will remove NM settings and add 'auto br0' settings elsewhere. Where did you
put 'auto br0, etc' into configuration file in /etc folder?
Tim
________________________________
From: John Forecast <john(a)forecast.name>
Sent: Wednesday, October 25, 2023 8:30 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
Bridges seem to interact badly with systemd. systemd does not wait for the bridge to
come up, it only waits for the physical ethernet device.
The result is that when DECnet tries to change the MAC address of br0 it does not exist.
The simplest way to make this work is to set the bridge’s MAC address when you create it.
Here’s my configuration file for creating br0:
auto br0
iface br0 inet static
bridge_hw aa:00:04:00:ca:a4
address 192.168.1.39/24
gateway 192.168.1.1
bridge_ports eth0 tap0
bridge_stp off
bridge_maxwait 5
dns-nameservers 192.168.1.12
Yours should read:
bridge_hw aa:00:04:00:9a:7c
If you use explicit brctl commands to set up your bridge, you’ll have to add the
following command after initially creating br0:
ip link set br0 address aa:00:04:00:9a:7c
John.
On Oct 25, 2023, at 7:53 PM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
John,
I already did that. It still did not work.
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line br0
node 31.153 name TSRTR
Tim
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 7:35 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
What does /etc/decnet.conf look like? In order to rebuild the configuration files you need
to answer “1” to the following question:
There appears to be an existing DECnet configuration present
Do you want to:
1 - Delete the existing configuration files and create new ones
2 - Use the existing configuration files
Enter code (1 - 2):
otherwise it will reuse the old configuration files.
John.
On Oct 25, 2023, at 6:59 PM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
No, it did not work. TSRTR and VAX emulators can't see HECnet world with enp4s0. I
tried with br0 and it worked fine. I do not need tap0 anymore.
I tried to change to br0 by re-build configuration files with Linux DECnet. decnet3
crashed when it attempted to connect br0 interface. It said node_address is missing.
I created br0 with using brctl commands. When I upgraded Ubuntu systems for some years,
it preserved br0 configurations.
Tim
I changed a line in tsrtr.conf from pcap:tap0 line:
circuit eth-0 Ethernet pcap:br0
$ systemctl status decnet3
× decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: failed (Result: exit-code) since Wed 2023-10-25 18:29:05 EDT; 58s >
Process: 1097 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1140 ExecStart=/usr/local/sbin/dnetd (code=exited, status=255/EXCE>
Main PID: 1140 (code=exited, status=255/EXCEPTION)
CPU: 43ms
Oct 25 18:29:05 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 18:29:05 tsux1a dnetd[1140]: /proc/sys/net/decnet/node_address missing, >
Oct 25 18:29:05 tsux1a systemd[1]: decnet3.service: Main process exited, code=e>
Oct 25 18:29:05 tsux1a systemd[1]: decnet3.service: Failed with result 'exit-co>
Oct 25 18:29:05 tsux1a systemd[1]: Failed to start decnet3.service - Load DECne>
$ dneigh
Error: can not open DECnet neigh file: /proc/net/decnet_neigh: No such file or directory
$ brctl show
bridge name bridge id STP enabled interfaces
br0 8000.3a221a4cea5a yes enp4s0
virbr0 8000.525400766173 yes
$ cat /proc/net/decnet_dev
Name Flags T1 Timer1 T3 Timer3 T4 Timer4 BlkSize Pri State DevType Router
Peer
lo B 0000 0000 0000 0000 0000 0000 65533 000 00 loopback
br0 B 0000 0000 0000 0010 0000 0000 1498 000 00 ethernet
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 5:53 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
I’m not familiar with the FIOS router configuration. I’m going to guess that there’s
one or more ethernet’s coming out of it; one going to TSRTR
and another going to TSUX1A. On TSRTR that is probably “enp4s0”. You will need to add a
circuit to pydecnet so it does routing on your
ethernet segment, something like:
circuit eth-0 Ethernet pcap:enp4s0
John.
On Oct 25, 2023, at 5:29 PM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
John,
I set virtual br0 for 8 tap devices that VAX emulators and others use. br0 is the main
routing. virbr0 is for QEMU VM (OpenVMS x86). TSUX1A is IP 192.168.1.228.
TSUX1A is Ubuntu 23.10 system on my embedded box. TSRTR is pydecnet router for HECnet
world access that connected to tap0. I am figuring out why TSRTR and TSUX1A can't see
each other. I am only able access TSVMSB (on SIMH VAX emulator) that uses tap1 through
TCP/IP but not DECnet.
TSUX1A is connected to my Verizon FiOS router for Internet access.
Here is ifconfig output below.
Tim
br0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 192.168.1.228 netmask 255.255.255.0 broadcast 192.168.1.255
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c prefixlen 64 scopeid
0x0<global>
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39 prefixlen 64 scopeid
0x0<global>
inet6 fe80::c288:756:b4bc:213e prefixlen 64 scopeid 0x20<link>
ether 3a:22:1a:4c:ea:5a txqueuelen 1000 (Ethernet)
RX packets 85254 bytes 10370418 (10.3 MB)
RX errors 0 dropped 8364 overruns 0 frame 0
TX packets 8096 bytes 813053 (813.0 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
enp4s0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
ether aa:00:04:00:9a:7c txqueuelen 1000 (Ethernet)
RX packets 100400 bytes 16878747 (16.8 MB)
RX errors 0 dropped 14 overruns 0 frame 0
TX packets 21289 bytes 2748243 (2.7 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
lo: flags=73<UP,LOOPBACK,RUNNING> mtu 65536
inet 127.0.0.1 netmask 255.0.0.0
inet6 ::1 prefixlen 128 scopeid 0x10<host>
loop txqueuelen 1000 (Local Loopback)
RX packets 1978 bytes 538618 (538.6 KB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 1978 bytes 538618 (538.6 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet6 fe80::884d:baff:fe98:9c38 prefixlen 64 scopeid 0x20<link>
ether 8a:4d:ba:98:9c:38 txqueuelen 1000 (Ethernet)
RX packets 5181 bytes 1525244 (1.5 MB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 74004 bytes 7346087 (7.3 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap1: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 6a:29:52:18:5a:14 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap2: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:90:0b:15:c2:7c txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap3: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 86:ae:14:ac:75:27 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap4: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether e6:d7:b6:a4:6e:2b txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap5: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 12:2f:e4:99:a2:70 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap6: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:21:92:ad:88:a0 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap7: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 9e:fd:b7:cd:a3:e6 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
virbr0: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
inet 192.168.122.1 netmask 255.255.255.0 broadcast 192.168.122.255
ether 52:54:00:76:61:73 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
ip addr:
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default
qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: enp4s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether aa:00:04:00:9a:7c brd ff:ff:ff:ff:ff:ff permaddr 00:c0:08:97:1a:2a
3: wlp3s0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen
1000
link/ether 54:14:f3:52:56:9d brd ff:ff:ff:ff:ff:ff
4: br0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group
default qlen 1000
link/ether 3a:22:1a:4c:ea:5a brd ff:ff:ff:ff:ff:ff
inet 192.168.1.228/24 brd 192.168.1.255 scope global dynamic noprefixroute br0
valid_lft 51908sec preferred_lft 51908sec
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c/64 scope global temporary dynamic
valid_lft 7193sec preferred_lft 7193sec
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39/64 scope global dynamic mngtmpaddr
noprefixroute
valid_lft 7193sec preferred_lft 7193sec
inet6 fe80::c288:756:b4bc:213e/64 scope link noprefixroute
valid_lft forever preferred_lft forever
5: virbr0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN
group default qlen 1000
link/ether 52:54:00:76:61:73 brd ff:ff:ff:ff:ff:ff
inet 192.168.122.1/24 brd 192.168.122.255 scope global virbr0
valid_lft forever preferred_lft forever
6: tap0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether 8a:4d:ba:98:9c:38 brd ff:ff:ff:ff:ff:ff
inet6 fe80::884d:baff:fe98:9c38/64 scope link
valid_lft forever preferred_lft forever
7: tap1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 6a:29:52:18:5a:14 brd ff:ff:ff:ff:ff:ff
8: tap2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:90:0b:15:c2:7c brd ff:ff:ff:ff:ff:ff
9: tap3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 86:ae:14:ac:75:27 brd ff:ff:ff:ff:ff:ff
10: tap4: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether e6:d7:b6:a4:6e:2b brd ff:ff:ff:ff:ff:ff
11: tap5: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 12:2f:e4:99:a2:70 brd ff:ff:ff:ff:ff:ff
12: tap6: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:21:92:ad:88:a0 brd ff:ff:ff:ff:ff:ff
13: tap7: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 9e:fd:b7:cd:a3:e6 brd ff:ff:ff:ff:ff:ff
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 5:05 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
What does your physical topology look like? You have TSRTR with routing between an
incoming Multinet connection and tap0, which is a
software construct so your VAX emulators can connect to HECnet. What is TSUX1A and how is
it connected to TSRTR? The output of “ifconfig”
or “ip addr” would probably be useful here - executed on both TSRTR and TSUX1A.
John.
On Oct 25, 2023, at 3:41 PM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
Here is my tsrtr.conf for pydecnet
circuit mul-0 Multinet <hecnet gw address>:connect
circuit tap-0 Ethernet tap:tap0
routing 31.153 --type l1router --bct1=20
nsp --qmax=5
node 31.153 TSRTR
node @nodenames.dat
system --ident "TSRTR PyDECnet router (Ubuntu system)"
I did not enable web interface into TSRTR server. I need to add circuit to connect an
end-node decnet for TSUX1A.
I have br0 and enp4s0 interface. Which interface should be added so that TSUX1A can see
TSRTR router?
I have tap-0 line so that VAX emulators can see HECnet world.
Tim
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 3:08 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
On Oct 25, 2023, at 1:43 PM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
John,
Ok. I first thought MIM:: is 1.1::. I now mean 1.13::
systemctl status decnet3
● decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: active (exited) since Wed 2023-10-25 13:24:51 EDT; 1min 2s ago
Process: 1089 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1122 ExecStart=/usr/local/sbin/dnetd (code=exited, status=0/SUCCES>
Main PID: 1122 (code=exited, status=0/SUCCESS)
Tasks: 1 (limit: 37150)
Memory: 4.1M
CPU: 47ms
CGroup: /system.slice/decnet3.service
└─1133 /usr/local/sbin/dnetd
Oct 25 13:24:51 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 13:24:51 tsux1a systemd[1]: Finished decnet3.service - Load DECnet modul>
That looks good.
dneigh
Node HWtype HWaddress Flags MTU Iface
TSUX1A loop AA:00:04:00:9A:7C --- 65533 lo
That also looks good, although there should be a second line with the designated router.
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line enp4s0
OK.
I tried to add tsrtr to decnet.conf but dneigh showed that it routes to enp4s0 instead of
br0.
This release is an end-node only implementation. There can only be a single “executor”
line which defines the local name, address and
physical (ethernet or wifi) device. All other lines define node address/name pairs. There
is no concept of setting a route to a node like in TCP/IP.
TSUX1A is expected to learn about TSRTR via a multicast message on the LAN connecting
TSRTR and TSUX1A.
I tried to configure br0 instead of enp4s0 but it crashed during booting. systemctl
showed that loading failure with attempting connect to br0 interface.
Yes, that seems reasonable.
It still can't see my pydecnet router (tsrtr 31.153).
Do you know how to configure decnet with br0 interface? How to set up a route to TSRTR
(31.153)?
I think the main problem here is that TSUX1A is not seeing TSRTR. How are they connected?
What does the WEB interface into TSRTR show?
John.
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 12:17 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
On Oct 25, 2023, at 10:47 AM, Timothy Stark
<fsword007@gmail.com<mailto:fsword007@gmail.com>> wrote:
I successfully installed Linux DECnet (beta test) on Ubuntu 23.10 but have some problems
with that.
I executed 'sudo ./BuildAndInstall.sh' and followed instructions to enter, etc.
Also I started pydecnet (route to HECnet network) as bridge.
I tried to set host to my VAX simulator but got no route to host errors.
Also I tried to access 1.1:: but got no route host errors as well. Look that below.
dndir 1.1::
connect failed: No route to host
I see this as well.
sethost tsvmsb
sethost V1.0.4
Connecting to tsvmsb
connect (cterm): No route to host
ct_setup_link: error 0 connecting to host
I don’t have a VMS system set up to try this but I get further trying to connect to
RSX-11m+:
forecast@emulat:~$ sethost 41.200
sethost V1.0.4
Connecting to 41.200
ct_setup_link: error sending init sequence
sudo ncp show known nodes
Version mismatch
You don’t need the “sudo” for normal operation. The “Version mismatch” is in response to
the connection completing with the wrong
optional data. Did this system have the previous DECnet for Linux installed? If so, did
you follow the section in README.DECnet
about deleting old libraries.
sudo ncp copy known nodes from 1.1
Unable to connect to listener
Same problem as dndir above.
lsmod | grep decnet3
decnet3 86016 5
That looks good. What about the output from “systemctl status decnet3”
John.
Do you have any solutions with those problems? On my VAX simulator (TSVMSB), I was able
access HECnet without any problems.
I tried to set host to Ubuntu 23.10 but immediately got error - network is unreachable.
Tim
________________________________
From: John Forecast <john@forecast.name<mailto:john@forecast.name>>
Sent: Tuesday, October 24, 2023 10:50 PM
To: The Hobbyist DECnet mailing list
<hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to
rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was
alway marked as
“experimental” and I never tried to get it running. If you need a DECnet
router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux
versions
were changes to the kernel framework APIs. By limiting the use of these
APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel
changes. So far,
I have had to make one additional conditional code change at kernel 6.5
and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal” is
still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management listener.
- There is now a subset implementation of “ncp” which is sufficient to support all
of the requests
available from the Linux network management listener. It does implement a “tell”
prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node
name database.
- The new kernel module now supports node counters as defined in the Network
Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi
releases. The current
procedure supports installation on systems derived from Debian or Fedora
although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet@lists.dfupdate.se<mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to
hecnet-leave@lists.dfupdate.se<mailto:hecnet-leave@lists.dfupdate.se>
3:53 p.m.
On Oct 26, 2023, at 7:51 AM, Timothy Stark
<fsword007(a)gmail.com> wrote:
Ok, I now got it. I use NetworkManager (nmcli commands) for adding br0 because I am
using Ubuntu system. I was searching for it and found configuration files in /etc/netplan
directory. I was looking for some manual about setting hwaddr but can't find it.
I tried to ip link set br0 address but it returned back to own addr after boot. decnet3
still failed to load.
That’s normal. You would have to change the MAC address after each boot, that’s what the
DECnetMAC systemd service does. Because bridges are purely software entities, systemd
doesn’t know anything about them and DECnetMAC runs before the bridge has been created.
I will remove NM settings and add 'auto br0'
settings elsewhere. Where did you put 'auto br0, etc' into configuration file in
/etc folder?
/etc/network/interfaces or as a separate file in /etc/network/interfaces.d
Ubuntu 32.10 does not include this file by default, you will need to install package
“ifupdown” and, probably, “net-tools”. I don’t know how this will interact with
NetworkManager.
John.
Tim
From: John Forecast <john(a)forecast.name>
Sent: Wednesday, October 25, 2023 8:30 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
Bridges seem to interact badly with systemd. systemd does not wait for the bridge to
come up, it only waits for the physical ethernet device.
The result is that when DECnet tries to change the MAC address of br0 it does not exist.
The simplest way to make this work is to set the bridge’s MAC address when you create it.
Here’s my configuration file for creating br0:
auto br0
iface br0 inet static
bridge_hw aa:00:04:00:ca:a4
address 192.168.1.39/24
gateway 192.168.1.1
bridge_ports eth0 tap0
bridge_stp off
bridge_maxwait 5
dns-nameservers 192.168.1.12
Yours should read:
bridge_hw aa:00:04:00:9a:7c
If you use explicit brctl commands to set up your bridge, you’ll have to add the
following command after initially creating br0:
ip link set br0 address aa:00:04:00:9a:7c
John.
On Oct 25, 2023, at 7:53 PM, Timothy Stark
<fsword007(a)gmail.com <mailto:fsword007@gmail.com>> wrote:
John,
I already did that. It still did not work.
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line br0
node 31.153 name TSRTR
Tim
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 7:35 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
What does /etc/decnet.conf look like? In order to rebuild the configuration files you
need to answer “1” to the following question:
There appears to be an existing DECnet configuration present
Do you want to:
1 - Delete the existing configuration files and create new ones
2 - Use the existing configuration files
Enter code (1 - 2):
otherwise it will reuse the old configuration files.
John.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se On Oct 25, 2023, at 6:59 PM, Timothy Stark
<fsword007(a)gmail.com <mailto:fsword007@gmail.com>> wrote:
No, it did not work. TSRTR and VAX emulators can't see HECnet world with enp4s0. I
tried with br0 and it worked fine. I do not need tap0 anymore.
I tried to change to br0 by re-build configuration files with Linux DECnet. decnet3
crashed when it attempted to connect br0 interface. It said node_address is missing.
I created br0 with using brctl commands. When I upgraded Ubuntu systems for some years,
it preserved br0 configurations.
Tim
I changed a line in tsrtr.conf from pcap:tap0 line:
circuit eth-0 Ethernet pcap:br0
$ systemctl status decnet3
× decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: failed (Result: exit-code) since Wed 2023-10-25 18:29:05 EDT; 58s >
Process: 1097 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1140 ExecStart=/usr/local/sbin/dnetd (code=exited, status=255/EXCE>
Main PID: 1140 (code=exited, status=255/EXCEPTION)
CPU: 43ms
Oct 25 18:29:05 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 18:29:05 tsux1a dnetd[1140]: /proc/sys/net/decnet/node_address missing, >
Oct 25 18:29:05 tsux1a systemd[1]: decnet3.service: Main process exited, code=e>
Oct 25 18:29:05 tsux1a systemd[1]: decnet3.service: Failed with result 'exit-co>
Oct 25 18:29:05 tsux1a systemd[1]: Failed to start decnet3.service - Load DECne>
$ dneigh
Error: can not open DECnet neigh file: /proc/net/decnet_neigh: No such file or directory
$ brctl show
bridge name bridge id STP enabled interfaces
br0 8000.3a221a4cea5a yes enp4s0
virbr0 8000.525400766173 yes
$ cat /proc/net/decnet_dev
Name Flags T1 Timer1 T3 Timer3 T4 Timer4 BlkSize Pri State DevType Router
Peer
lo B 0000 0000 0000 0000 0000 0000 65533 000 00 loopback
br0 B 0000 0000 0000 0010 0000 0000 1498 000 00 ethernet
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 5:53 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
I’m not familiar with the FIOS router configuration. I’m going to guess that there’s
one or more ethernet’s coming out of it; one going to TSRTR
and another going to TSUX1A. On TSRTR that is probably “enp4s0”. You will need to add a
circuit to pydecnet so it does routing on your
ethernet segment, something like:
circuit eth-0 Ethernet pcap:enp4s0
John.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se> On Oct 25, 2023, at 5:29 PM, Timothy Stark
<fsword007(a)gmail.com <mailto:fsword007@gmail.com>> wrote:
John,
I set virtual br0 for 8 tap devices that VAX emulators and others use. br0 is the main
routing. virbr0 is for QEMU VM (OpenVMS x86). TSUX1A is IP 192.168.1.228.
TSUX1A is Ubuntu 23.10 system on my embedded box. TSRTR is pydecnet router for HECnet
world access that connected to tap0. I am figuring out why TSRTR and TSUX1A can't see
each other. I am only able access TSVMSB (on SIMH VAX emulator) that uses tap1 through
TCP/IP but not DECnet.
TSUX1A is connected to my Verizon FiOS router for Internet access.
Here is ifconfig output below.
Tim
br0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 192.168.1.228 netmask 255.255.255.0 broadcast 192.168.1.255
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c prefixlen 64 scopeid
0x0<global>
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39 prefixlen 64 scopeid
0x0<global>
inet6 fe80::c288:756:b4bc:213e prefixlen 64 scopeid 0x20<link>
ether 3a:22:1a:4c:ea:5a txqueuelen 1000 (Ethernet)
RX packets 85254 bytes 10370418 (10.3 MB)
RX errors 0 dropped 8364 overruns 0 frame 0
TX packets 8096 bytes 813053 (813.0 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
enp4s0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
ether aa:00:04:00:9a:7c txqueuelen 1000 (Ethernet)
RX packets 100400 bytes 16878747 (16.8 MB)
RX errors 0 dropped 14 overruns 0 frame 0
TX packets 21289 bytes 2748243 (2.7 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
lo: flags=73<UP,LOOPBACK,RUNNING> mtu 65536
inet 127.0.0.1 netmask 255.0.0.0
inet6 ::1 prefixlen 128 scopeid 0x10<host>
loop txqueuelen 1000 (Local Loopback)
RX packets 1978 bytes 538618 (538.6 KB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 1978 bytes 538618 (538.6 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet6 fe80::884d:baff:fe98:9c38 prefixlen 64 scopeid 0x20<link>
ether 8a:4d:ba:98:9c:38 txqueuelen 1000 (Ethernet)
RX packets 5181 bytes 1525244 (1.5 MB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 74004 bytes 7346087 (7.3 MB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap1: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 6a:29:52:18:5a:14 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap2: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:90:0b:15:c2:7c txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap3: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 86:ae:14:ac:75:27 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap4: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether e6:d7:b6:a4:6e:2b txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap5: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 12:2f:e4:99:a2:70 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap6: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 82:21:92:ad:88:a0 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
tap7: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether 9e:fd:b7:cd:a3:e6 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
virbr0: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
inet 192.168.122.1 netmask 255.255.255.0 broadcast 192.168.122.255
ether 52:54:00:76:61:73 txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
ip addr:
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default
qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: enp4s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether aa:00:04:00:9a:7c brd ff:ff:ff:ff:ff:ff permaddr 00:c0:08:97:1a:2a
3: wlp3s0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen
1000
link/ether 54:14:f3:52:56:9d brd ff:ff:ff:ff:ff:ff
4: br0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group
default qlen 1000
link/ether 3a:22:1a:4c:ea:5a brd ff:ff:ff:ff:ff:ff
inet 192.168.1.228/24 brd 192.168.1.255 scope global dynamic noprefixroute br0
valid_lft 51908sec preferred_lft 51908sec
inet6 2600:4040:2cb8:ac00:4d41:6e91:93c0:e35c/64 scope global temporary dynamic
valid_lft 7193sec preferred_lft 7193sec
inet6 2600:4040:2cb8:ac00:3dbd:4b54:a857:4e39/64 scope global dynamic mngtmpaddr
noprefixroute
valid_lft 7193sec preferred_lft 7193sec
inet6 fe80::c288:756:b4bc:213e/64 scope link noprefixroute
valid_lft forever preferred_lft forever
5: virbr0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN
group default qlen 1000
link/ether 52:54:00:76:61:73 brd ff:ff:ff:ff:ff:ff
inet 192.168.122.1/24 brd 192.168.122.255 scope global virbr0
valid_lft forever preferred_lft forever
6: tap0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master br0
state UP group default qlen 1000
link/ether 8a:4d:ba:98:9c:38 brd ff:ff:ff:ff:ff:ff
inet6 fe80::884d:baff:fe98:9c38/64 scope link
valid_lft forever preferred_lft forever
7: tap1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 6a:29:52:18:5a:14 brd ff:ff:ff:ff:ff:ff
8: tap2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:90:0b:15:c2:7c brd ff:ff:ff:ff:ff:ff
9: tap3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 86:ae:14:ac:75:27 brd ff:ff:ff:ff:ff:ff
10: tap4: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether e6:d7:b6:a4:6e:2b brd ff:ff:ff:ff:ff:ff
11: tap5: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 12:2f:e4:99:a2:70 brd ff:ff:ff:ff:ff:ff
12: tap6: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 82:21:92:ad:88:a0 brd ff:ff:ff:ff:ff:ff
13: tap7: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast master br0
state DOWN group default qlen 1000
link/ether 9e:fd:b7:cd:a3:e6 brd ff:ff:ff:ff:ff:ff
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 5:05 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
What does your physical topology look like? You have TSRTR with routing between an
incoming Multinet connection and tap0, which is a
software construct so your VAX emulators can connect to HECnet. What is TSUX1A and how is
it connected to TSRTR? The output of “ifconfig”
or “ip addr” would probably be useful here - executed on both TSRTR and TSUX1A.
John.
On Oct 25, 2023, at 3:41 PM, Timothy Stark <fsword007(a)gmail.com
<mailto:fsword007@gmail.com>> wrote:
Here is my tsrtr.conf for pydecnet
circuit mul-0 Multinet <hecnet gw address>:connect
circuit tap-0 Ethernet tap:tap0
routing 31.153 --type l1router --bct1=20
nsp --qmax=5
node 31.153 TSRTR
node @nodenames.dat
system --ident "TSRTR PyDECnet router (Ubuntu system)"
I did not enable web interface into TSRTR server. I need to add circuit to connect an
end-node decnet for TSUX1A.
I have br0 and enp4s0 interface. Which interface should be added so that TSUX1A can see
TSRTR router?
I have tap-0 line so that VAX emulators can see HECnet world.
Tim
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 3:08 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Tim,
On Oct 25, 2023, at 1:43 PM, Timothy Stark <fsword007(a)gmail.com
<mailto:fsword007@gmail.com>> wrote:
John,
Ok. I first thought MIM:: is 1.1::. I now mean 1.13::
systemctl status decnet3
● decnet3.service - Load DECnet module and start
Loaded: loaded (/etc/systemd/system/decnet3.service; enabled; preset: enab>
Active: active (exited) since Wed 2023-10-25 13:24:51 EDT; 1min 2s ago
Process: 1089 ExecStartPre=/usr/local/sbin/dnetLoadModule (code=exited, sta>
Process: 1122 ExecStart=/usr/local/sbin/dnetd (code=exited, status=0/SUCCES>
Main PID: 1122 (code=exited, status=0/SUCCESS)
Tasks: 1 (limit: 37150)
Memory: 4.1M
CPU: 47ms
CGroup: /system.slice/decnet3.service
└─1133 /usr/local/sbin/dnetd
Oct 25 13:24:51 tsux1a systemd[1]: Starting decnet3.service - Load DECnet modul>
Oct 25 13:24:51 tsux1a systemd[1]: Finished decnet3.service - Load DECnet modul>
That looks good.
dneigh
Node HWtype HWaddress Flags MTU Iface
TSUX1A loop AA:00:04:00:9A:7C --- 65533 lo
That also looks good, although there should be a second line with the designated router.
cat /etc/decnet.conf
#V001.0
# DECnet hosts file
#
#Node Node Name Node Line Line
#Type Address Tag Name Tag Device
#----- ------- ----- ----- ----- ------
executor 31.154 name TSUX1A line enp4s0
OK.
I tried to add tsrtr to decnet.conf but dneigh showed that it routes to enp4s0 instead of
br0.
This release is an end-node only implementation. There can only be a single “executor”
line which defines the local name, address and
physical (ethernet or wifi) device. All other lines define node address/name pairs. There
is no concept of setting a route to a node like in TCP/IP.
TSUX1A is expected to learn about TSRTR via a multicast message on the LAN connecting
TSRTR and TSUX1A.
I tried to configure br0 instead of enp4s0 but it crashed during booting. systemctl
showed that loading failure with attempting connect to br0 interface.
Yes, that seems reasonable.
It still can't see my pydecnet router (tsrtr 31.153).
Do you know how to configure decnet with br0 interface? How to set up a route to TSRTR
(31.153)?
I think the main problem here is that TSUX1A is not seeing TSRTR. How are they connected?
What does the WEB interface into TSRTR show?
John.
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Wednesday, October 25, 2023 12:17 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Hi Tim,
I get the same result with “dndir 1.1.::” - is MAGICA down? I have no problems
accessing MIM (1.13).
What output do you get from “dneigh” - that should display your local system address and
designated router, here’s mine:
forecast@emulat:~$ dneigh
Node HWtype HWaddress Flags MTU Iface
EMULAT loop AA:00:04:00:CA:A4 --- 65533 lo
41.250 ether AA:00:04:00:FA:A4 1-- 591 br0
41.235 ether AA:00:04:00:EB:A4 --- 1498 br0
28NH ether AA:00:04:00:1C:A4 --- 596 br0
MIM ether AA:00:04:00:0D:04 --- 591 br0
MAGICA ether AA:00:04:00:01:04 --- 591 br0
PYTHON ether AA:00:04:00:01:A4 --- 591 br0
41.250 is 4CAST::, my router running pyDECnet.
On Oct 25, 2023, at 10:47 AM, Timothy Stark <fsword007(a)gmail.com
<mailto:fsword007@gmail.com>> wrote:
I successfully installed Linux DECnet (beta test) on Ubuntu 23.10 but have some problems
with that.
I executed 'sudo ./BuildAndInstall.sh' and followed instructions to enter, etc.
Also I started pydecnet (route to HECnet network) as bridge.
I tried to set host to my VAX simulator but got no route to host errors.
Also I tried to access 1.1:: but got no route host errors as well. Look that below.
dndir 1.1::
connect failed: No route to host
I see this as well.
sethost tsvmsb
sethost V1.0.4
Connecting to tsvmsb
connect (cterm): No route to host
ct_setup_link: error 0 connecting to host
I don’t have a VMS system set up to try this but I get further trying to connect to
RSX-11m+:
forecast@emulat:~$ sethost 41.200
sethost V1.0.4
Connecting to 41.200
ct_setup_link: error sending init sequence
sudo ncp show known nodes
Version mismatch
You don’t need the “sudo” for normal operation. The “Version mismatch” is in response to
the connection completing with the wrong
optional data. Did this system have the previous DECnet for Linux installed? If so, did
you follow the section in README.DECnet
about deleting old libraries.
sudo ncp copy known nodes from 1.1
Unable to connect to listener
Same problem as dndir above.
lsmod | grep decnet3
decnet3 86016 5
That looks good. What about the output from “systemctl status decnet3”
John.
Do you have any solutions with those problems? On my VAX simulator (TSVMSB), I was able
access HECnet without any problems.
I tried to set host to Ubuntu 23.10 but immediately got error - network is unreachable.
Tim
From: John Forecast <john(a)forecast.name <mailto:john@forecast.name>>
Sent: Tuesday, October 24, 2023 10:50 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se
<mailto:hecnet@lists.dfupdate.se>>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet
<https://github.com/JohnForecast/RaspbianDECnet>> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to
rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was
alway marked as
“experimental” and I never tried to get it running. If you need a DECnet
router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux
versions
were changes to the kernel framework APIs. By limiting the use of these
APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel
changes. So far,
I have had to make one additional conditional code change at kernel 6.5
and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal”
is still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management
listener.
- There is now a subset implementation of “ncp” which is sufficient to support
all of the requests
available from the Linux network management listener. It does implement a
“tell” prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node
name database.
- The new kernel module now supports node counters as defined in the Network
Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi
releases. The current
procedure supports installation on systems derived from Debian or Fedora
although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet
<https://github.com/JohnForecast/LinuxDECnet>>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
4:29 p.m.
On Oct 26, 2023, at 9:53 AM, John Forecast
<john(a)forecast.name> wrote:
Systemd supports (partial) ordering of operations, so presumably it would be possible to
add those ordering requirements to the bridge and decnet startup files to get things
happening in the right order.
paul
On Oct 26, 2023, at 7:51 AM, Timothy Stark <fsword007(a)gmail.com> wrote:
Ok, I now got it. I use NetworkManager (nmcli commands) for adding br0 because I am
using Ubuntu system. I was searching for it and found configuration files in /etc/netplan
directory. I was looking for some manual about setting hwaddr but can't find it.
I tried to ip link set br0 address but it returned back to own addr after boot. decnet3
still failed to load.
That’s normal. You would have to change the MAC address after each boot, that’s what the
DECnetMAC systemd service does. Because bridges are purely software entities, systemd
doesn’t know anything about them and DECnetMAC runs before the bridge has been created.
25 Oct
25 Oct
4:58 p.m.
On Oct 24, 2023, at 10:50 PM, John Forecast
<john(a)forecast.name> wrote:
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to rebuild the
entire
kernel. ...
Can this also be installed on older kernels where DECnet is still in the kernel source
tree, or does it only work for 6.1 and later?
My BeagleBone runs Debian, kernel 5.10.168, so can I try it there or should I build a
custom kernel with its DECnet enabled instead?
paul
6:22 p.m.
Hi Paul,
I’ve tested it on various x86 and Arm systems with kernels from 4.18.0 to 6.5.5. I
would not recommend upgrading a system running the old
DECnet for Linux code. Try to start with a new clean system.
John.
On Oct 25, 2023, at 10:58 AM, Paul Koning
<paulkoning(a)comcast.net> wrote:
On Oct 24, 2023, at 10:50 PM, John Forecast
<john(a)forecast.name> wrote:
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to rebuild the
entire
kernel. ...
Can this also be installed on older kernels where DECnet is still in the kernel source
tree, or does it only work for 6.1 and later?
My BeagleBone runs Debian, kernel 5.10.168, so can I try it there or should I build a
custom kernel with its DECnet enabled instead?
paul
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
6:28 p.m.
On Oct 25, 2023, at 12:22 PM, John Forecast
<john(a)forecast.name> wrote:
Hi Paul,
I’ve tested it on various x86 and Arm systems with kernels from 4.18.0 to 6.5.5. I
would not recommend upgrading a system running the old
DECnet for Linux code. Try to start with a new clean system.
John.
Thanks. My BBB runs a stock Debian for BBB release, so I'm testing your build script
now.
paul
6:12 p.m.
John,
Thank you for your work!
Finally a simple way to have a linux host with an up-to-date kernel (not just the pi:s)
I made the installed on a fresh Debian Bookworm
Two minor issues:
The link to the git site below seems to include the last "GT" sign which gives
the wrong URL to the site
The install script bails on adding additional hosts in the local db as below with an error
message and the continues.
Enter the node's address: area.node (e.g. 1.1) : 34.1023
Enter its node name : A34RTR
./BuildAndInstall.sh: command substitution: line 471: unexpected EOF while looking
for matching `"'
Other than that it seems to works just fine!
BR
/t
Den 2023-10-25 04:50 skrev "John Forecast" <john(a)forecast.name
<mailto:john@forecast.name>> följande:
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet>
<https://github.com/JohnForecast/RaspbianDECnet>> can no longer be installed
on recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to rebuild the
entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was alway marked as
“experimental” and I never tried to get it running. If you need a DECnet router, pyDECnet
or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux versions
were changes to the kernel framework APIs. By limiting the use of these APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel changes. So far,
I have had to make one additional conditional code change at kernel 6.5 and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal” is still
available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management listener.
- There is now a subset implementation of “ncp” which is sufficient to support all of the
requests
available from the Linux network management listener. It does implement a “tell” prefix so
all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node name
database.
- The new kernel module now supports node counters as defined in the Network Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi releases. The
current
procedure supports installation on systems derived from Debian or Fedora although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet>
<https://github.com/JohnForecast/LinuxDECnet>>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
6:38 p.m.
Hi Tomas,
Thanks for the bug report, I’ve updated the script in the repository.
John.
On Oct 25, 2023, at 12:12 PM, Tomas Prybil
<tomas(a)prybil.se> wrote:
John,
Thank you for your work!
Finally a simple way to have a linux host with an up-to-date kernel (not just the pi:s)
I made the installed on a fresh Debian Bookworm
Two minor issues:
The link to the git site below seems to include the last "GT" sign which gives
the wrong URL to the site
The install script bails on adding additional hosts in the local db as below with an
error message and the continues.
Enter the node's address: area.node (e.g. 1.1) : 34.1023
Enter its node name : A34RTR
./BuildAndInstall.sh: command substitution: line 471: unexpected EOF while looking
for matching `"'
Other than that it seems to works just fine!
BR
/t
Den 2023-10-25 04:50 skrev "John Forecast" <john(a)forecast.name
<mailto:john@forecast.name>> följande:
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet>
<https://github.com/JohnForecast/RaspbianDECnet>> can no longer be installed
on recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to rebuild the
entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was alway marked
as
“experimental” and I never tried to get it running. If you need a DECnet router, pyDECnet
or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux versions
were changes to the kernel framework APIs. By limiting the use of these APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel changes. So far,
I have had to make one additional conditional code change at kernel 6.5 and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal” is still
available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management listener.
- There is now a subset implementation of “ncp” which is sufficient to support all of the
requests
available from the Linux network management listener. It does implement a “tell” prefix
so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node name
database.
- The new kernel module now supports node counters as defined in the Network Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi releases. The
current
procedure supports installation on systems derived from Debian or Fedora although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet>
<https://github.com/JohnForecast/LinuxDECnet>>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se <mailto:hecnet@lists.dfupdate.se>
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
<mailto:hecnet-leave@lists.dfupdate.se>
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
9 p.m.
Thank you John! Brilliant work. I now have a DEBIAN 12 box on HECnet, in
a matter of minutes :-).
Granted I have not taken it for much of a test drive, but so far there
appear to be zero issues.
Mark
On 25/10/2023 03:50, John Forecast wrote:
Sometime around mid-2022 the Linux kernel developers
decided to remove the DECnet code from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to rebuild the
entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was alway marked
as
“experimental” and I never tried to get it running. If you need a DECnet router,
pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux versions
were changes to the kernel framework APIs. By limiting the use of these APIs I am
hoping
that the kernel module will need fewer changes to keep up with kernel changes. So far,
I have had to make one additional conditional code change at kernel 6.5 and that was
in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal” is still
available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management listener.
- There is now a subset implementation of “ncp” which is sufficient to support all of
the requests
available from the Linux network management listener. It does implement a “tell”
prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node name
database.
- The new kernel module now supports node counters as defined in the Network Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi releases. The
current
procedure supports installation on systems derived from Debian or Fedora although only
a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory
for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
26 Oct
26 Oct
4:31 a.m.
On Tue, Oct 24, 2023 at 10:50 PM John Forecast <john(a)forecast.name> wrote:
Sometime around mid-2022 the Linux kernel developers
decided to remove the DECnet code from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
John, forgive me if this question is obtuse, but I'm curious: is there
any reason this couldn't be hoisted entirely into userspace and built
using TAP/TUN, which would put a (relatively) clean interface between
DECnet and the kernel?
- Dan C.
4:59 a.m.
Hi Dan,
From the network receive/transmit side that would certainly be possible with minimal,
if any, kernel changes. Unfortunately you still have to worry about the application APIs
(connect(), accept(), read(), write() etc) and there is no easy way to hoist them into
userspace without substantial kernel changes. You would need something similar to FUSE.
John.
On Oct 25, 2023, at 10:31 PM, Dan Cross
<crossd(a)gmail.com> wrote:
On Tue, Oct 24, 2023 at 10:50 PM John Forecast <john(a)forecast.name> wrote:
Sometime around mid-2022 the Linux kernel
developers decided to remove the DECnet code from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
John, forgive me if this question is obtuse, but I'm curious: is there
any reason this couldn't be hoisted entirely into userspace and built
using TAP/TUN, which would put a (relatively) clean interface between
DECnet and the kernel?
- Dan C.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
1:43 p.m.
You can sign kernel driver by using kmodsign. I think that you could add new network
protocols to the kernel without re-compiling. Google it.
I Googled it and now learned some things about kernel development.
Tim
________________________________
From: John Forecast <john(a)forecast.name>
Sent: Wednesday, October 25, 2023 10:59 PM
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se>
Subject: [HECnet] Re: Announcing new release of Linux DECnet
Hi Dan,
From the network receive/transmit side that would certainly be possible with minimal,
if any, kernel changes. Unfortunately you still have to worry about the application APIs
(connect(), accept(), read(), write() etc) and there is no easy way to hoist them into
userspace without substantial kernel changes. You would need something similar to FUSE.
John.
On Oct 25, 2023, at 10:31 PM, Dan Cross
<crossd(a)gmail.com> wrote:
On Tue, Oct 24, 2023 at 10:50 PM John Forecast <john(a)forecast.name> wrote:
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
Sometime around mid-2022 the Linux kernel
developers decided to remove the DECnet code from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
John, forgive me if this question is obtuse, but I'm curious: is there
any reason this couldn't be hoisted entirely into userspace and built
using TAP/TUN, which would put a (relatively) clean interface between
DECnet and the kernel?
- Dan C.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
30 Oct
30 Oct
1:10 a.m.
On Wed, Oct 25, 2023 at 11:00 PM John Forecast <john(a)forecast.name> wrote:
Hi Dan,
From the network receive/transmit side that would certainly be possible with minimal,
if any, kernel changes. Unfortunately you still have to worry about the application APIs
(connect(), accept(), read(), write() etc) and there is no easy way to hoist them into
userspace without substantial kernel changes. You would need something similar to FUSE.
Hi John, thanks for the note back.
For the application-side system calls, I suppose I thought that one
might fake most of it by clever use of LD_PRELOAD or the like to
overload socket(), connect(), et al, and redirect those to an RPC-ish
protocol over a Unix domain socket to the DECnet server. But that
obviously won't support existing binaries that make raw system calls
(as opposed to invoking stubs in the C library).
- Dan C.
On Oct 25,
2023, at 10:31 PM, Dan Cross <crossd(a)gmail.com> wrote:
On Tue, Oct 24, 2023 at 10:50 PM John Forecast <john(a)forecast.name> wrote:
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se Sometime around mid-2022 the Linux kernel
developers decided to remove the DECnet code from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
John, forgive me if this question is obtuse, but I'm curious: is there
any reason this couldn't be hoisted entirely into userspace and built
using TAP/TUN, which would put a (relatively) clean interface between
DECnet and the kernel?
- Dan C.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
26 Oct
26 Oct
4:27 p.m.
On Oct 25, 2023, at 10:31 PM, Dan Cross
<crossd(a)gmail.com> wrote:
On Tue, Oct 24, 2023 at 10:50 PM John Forecast <john(a)forecast.name> wrote:
That would deal with the network interface end, but the other end (the API) is sockets,
and I don't think there is a FUSE analog for sockets that lets you implement a socket
address family in user mode. If there were, that might be enough to do what you
describe.
This is why the PyDECnet API is clunky; it needs to handle requests relating to one or
more full duplex data connections, but it's hard to find an elegant POSIX API that
PyDECnet can hook into. The least ugly I have found so far is the one in the T1.1 beta,
which uses pipes or Unix domain sockets to pass DECnet API requests encoded in JSON. It
works, but it certainly isn't transparent the way the DECnet/Linux socket extensions
are.
paul
Sometime around mid-2022 the Linux kernel
developers decided to remove the DECnet code from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
John, forgive me if this question is obtuse, but I'm curious: is there
any reason this couldn't be hoisted entirely into userspace and built
using TAP/TUN, which would put a (relatively) clean interface between
DECnet and the kernel?
6:31 p.m.
On Oct 26, 2023, at 10:27 AM, Paul Koning
<paulkoning(a)comcast.net> wrote:
As far as I know there is no FUSE like interface for sockets. Over the years I’ve seen at
least a couple of attempts started , but none make it past the early prototyping stage.
John.
On Oct 25, 2023, at 10:31 PM, Dan Cross
<crossd(a)gmail.com> wrote:
On Tue, Oct 24, 2023 at 10:50 PM John Forecast <john(a)forecast.name> wrote:
That would deal with the network interface end, but the other end (the API) is sockets,
and I don't think there is a FUSE analog for sockets that lets you implement a socket
address family in user mode. If there were, that might be enough to do what you
describe.
This is why the PyDECnet API is clunky; it needs to handle requests relating to one or
more full duplex data connections, but it's hard to find an elegant POSIX API that
PyDECnet can hook into. The least ugly I have found so far is the one in the T1.1 beta,
which uses pipes or Unix domain sockets to pass DECnet API requests encoded in JSON. It
works, but it certainly isn't transparent the way the DECnet/Linux socket extensions
are.
paul
Sometime around mid-2022 the Linux kernel
developers decided to remove the DECnet code from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
John, forgive me if this question is obtuse, but I'm curious: is there
any reason this couldn't be hoisted entirely into userspace and built
using TAP/TUN, which would put a (relatively) clean interface between
DECnet and the kernel? _______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
4:55 a.m.
On 10/24/23 22:50, John Forecast wrote:
Sometime around mid-2022 the Linux kernel developers
decided to remove the DECnet code from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
...
This is amazing! Thank you SO MUCH for doing this! I'll be trying
this out at LSSM as soon as I can.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
6:46 p.m.
On October 24, 2023 10:51:03 PM John Forecast <john(a)forecast.name> wrote:
Sometime around mid-2022 the Linux kernel developers
decided to remove the DECnet code from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet
implementation. More
recently, there has been some discussion around removing the code from the Long Term
Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
...
I took a few minutes to look at this. I was able to get it running at a basic level in
an Ubuntu 22.04 VM under SmartOS. No tweaks were required. I was able to log in to an
Itanium2 machine running VMS 8.3, and do a directory of MIM::. Everything worked as
expected in a brief test.
I will spend more time with it in the coming days.
Once again, I wish to express my sincere thanks for doing this.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
31 Oct
31 Oct
4:39 p.m.
Hi,
Thanks for doing all of this work. I have it running on an existing raspberry pi 4 32bit
OS having done a dist and kernel upgrade to 6.1.10...
I just need to change the service order startup so that it correctly uses a tap device and
changes its mac address, rather than that of the bridge.
And then I got rather a shock with something called NetworkManager that appeared when I
installed a new 64 bit version of Raspberry Pi OS.
Nmcli etc...! Perhaps I'm just getting old but why do people make this all so complex
with layer upon layer upon layer of utter garbage. That way lies Windows.
It's there for all to see! A huge mistake. And they wonder why the devious find so
many illicit routes into a host!
Anyway, rant over. I can disable NetworkManager with systemctl but it's also nice
enough to back off and not interfere with any interfaces referenced in
/etc/network/interfaces which is where I put all my bridge and tap stuff.
Keith
-----Original Message-----
From: John Forecast [mailto:john@forecast.name]
Sent: 25 October 2023 03:50
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux kernel. Kernel 6.0.x was the last release which included source code for a
DECnet implementation. More recently, there has been some discussion around removing the
code from the Long Term Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to rebuild the
entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was alway marked
as
“experimental” and I never tried to get it running. If you need a DECnet router,
pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux versions
were changes to the kernel framework APIs. By limiting the use of these APIs I am
hoping
that the kernel module will need fewer changes to keep up with kernel changes. So far,
I have had to make one additional conditional code change at kernel 6.5 and that was
in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal” is still
available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management listener.
- There is now a subset implementation of “ncp” which is sufficient to support all of the
requests
available from the Linux network management listener. It does implement a “tell” prefix
so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node name
database.
- The new kernel module now supports node counters as defined in the Network Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi releases. The
current
procedure supports installation on systems derived from Debian or Fedora although only
a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se To unsubscribe send an email to
hecnet-leave(a)lists.dfupdate.se
6:07 p.m.
Hi Keith,
You have to be careful with NetworkManager and defining device information in
/etc/network/interfaces - NetworkManager will back off but you also lose the startup
synchronization provided by systemd. I recently added a note to README.DECnet about this.
The basic idea is to define the MAC address of the bridge as part of the setup (this would
be the AA:00:04:00:xx:yy DECnet address) using:
bridge_hw=aa:00:04:00:xx:yy
and then you can disable DECnetMAC with systemctl if you don’t want to see an error
message during startup.
John.
On Oct 31, 2023, at 11:39 AM, Keith Halewood
<Keith.Halewood(a)pitbulluk.org> wrote:
Hi,
Thanks for doing all of this work. I have it running on an existing raspberry pi 4 32bit
OS having done a dist and kernel upgrade to 6.1.10...
I just need to change the service order startup so that it correctly uses a tap device
and changes its mac address, rather than that of the bridge.
And then I got rather a shock with something called NetworkManager that appeared when I
installed a new 64 bit version of Raspberry Pi OS.
Nmcli etc...! Perhaps I'm just getting old but why do people make this all so complex
with layer upon layer upon layer of utter garbage. That way lies Windows.
It's there for all to see! A huge mistake. And they wonder why the devious find so
many illicit routes into a host!
Anyway, rant over. I can disable NetworkManager with systemctl but it's also nice
enough to back off and not interfere with any interfaces referenced in
/etc/network/interfaces which is where I put all my bridge and tap stuff.
Keith
-----Original Message-----
From: John Forecast [mailto:john@forecast.name]
Sent: 25 October 2023 03:50
To: The Hobbyist DECnet mailing list <hecnet(a)lists.dfupdate.se>
Subject: [HECnet] Announcing new release of Linux DECnet
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code
from the Linux kernel. Kernel 6.0.x was the last release which included source code for a
DECnet implementation. More recently, there has been some discussion around removing the
code from the Long Term Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on
recent releases.
Over the past year or so, I have been working on a replacement for this repository with
the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to rebuild the
entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was alway marked
as
“experimental” and I never tried to get it running. If you need a DECnet router,
pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux versions
were changes to the kernel framework APIs. By limiting the use of these APIs I am
hoping
that the kernel module will need fewer changes to keep up with kernel changes. So far,
I have had to make one additional conditional code change at kernel 6.5 and that was
in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland
code:
- What used to be “fal2” is now the default file access listener. The old “fal” is still
available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management listener.
- There is now a subset implementation of “ncp” which is sufficient to support all of
the requests
available from the Linux network management listener. It does implement a “tell”
prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node name
database.
- The new kernel module now supports node counters as defined in the Network Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi releases. The
current
procedure supports installation on systems derived from Debian or Fedora although only
a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet>
and should be considered to be in Beta test at this time. See README.DECnet in the top
level directory for more information and installation instructions.
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se To unsubscribe send an email to
hecnet-leave(a)lists.dfupdate.se
_______________________________________________
HECnet mailing list -- hecnet(a)lists.dfupdate.se
To unsubscribe send an email to hecnet-leave(a)lists.dfupdate.se
191
days inactive
197
days old
37 comments
9 participants
participants (9)
-
Crni Mrki -
Dan Cross -
Dave McGuire -
John Forecast -
Keith Halewood -
Mark Curtis -
Paul Koning -
Timothy Stark -
Tomas Prybil