Oleg Safiullin wrote:
Weird. I tried setting host from MIM to MIM, loggin in as guest. No errors.
No idea what the problem could be...?
Something related to connection speed and multiprocessing? :)
btw, i tried to connect from MIM to CTAKAH and found no errors.
I'll try to do more tests...
Hmm.
No problems when i log in using my account...
--
>SET DCL
$ SH TER
RT1: [SAFIULLIN] [202,155] 12-NOV-11 17:09 1 O. SAFIULLIN
CLI = DCL BUF = 80. HFILL = 0
LINES = 24. TERM = VT2xx OWNER = none NOPARITY
NOPRINTER_PORT NOPASTHRU NOSERIAL
NOLOWER NOPRIV NOHOLD NOSLAVE ESC CRT NOFORM REMOTE
ECHO NOVFILL HHT NOFDX WRAP NORPA EBC NOTYPEAHEAD
NOCTRLC AVO ANSI DEC EDIT NOREGIS SOFT NOBLKMOD
HSYNC BRO NOABAUD TTSYNC
$ PIP
PIP>^Z
$ T EDTINI.EDT
DEFINE KEY GOLD L AS "SHL."
DEFINE KEY GOLD R AS "SHR."
SET MODE CHANGE
$
Weird. I tried setting host from MIM to MIM, loggin in as guest. No errors.
No idea what the problem could be...?
Something related to connection speed and multiprocessing? :)
btw, i tried to connect from MIM to CTAKAH and found no errors.
I'll try to do more tests...
On 2011-11-12 16.53, Oleg Safiullin wrote:
Johnny Billquist wrote:
On 2011-11-12 16.41, Oleg Safiullin wrote:
Huh. At what point? That should not happen. I don't see anything like
that when I try to log in as guest...
Just got the same error when i tried to log in as guest using SET HOST.
No problems for telnet login.
[...]
Oh! Darn! I bet I know what the problem is with that... The normal
terminal driver only requires read access to the buffer that is used
to write to the terminal. The RT: device driver, which is what is used
for the CTERM, probably wants write access, even though it will only
read from the buffer.
And then PIP (which is what TYPE invokes) does a QIO from a memory
buffer that is read only... Not sure I can do anything to fix that... :-(
I'll test and check some more...
(I consider the write access check a bug in the device driver, and
I've stumbled on it before when writing my own code...)
Johnny
Looks like even GCML doesn't work...
--
$ pip
PIP>
PIP -- Get Command Line - I/O Error
$ flx
FLX>
FLX -- I/O error on command input
$ run $mac
RT0>
RT0 -- Command I/O error
$
--
No problems on CTAKAH.
--
$ sh ter
RT0: [GUEST] [377,377] 12-NOV-11 22:50 1 A. GUEST
CLI = DCL BUF = 80. HFILL = 0
LINES = 24. TERM = VT2xx OWNER = RT0: NOPARITY
NOPRINTER_PORT NOPASTHRU NOSERIAL
LOWER NOPRIV NOHOLD NOSLAVE ESC CRT NOFORM REMOTE
NOECHO NOVFILL HHT NOFDX WRAP NORPA EBC NOTYPEAHEAD
NOCTRLC AVO ANSI DEC EDIT NOREGIS SOFT NOBLKMOD
HSYNC BRO NOABAUD TTSYNC
$ pip
PIP>^Z
$
--
Weird. I tried setting host from MIM to MIM, loggin in as guest. No errors.
No idea what the problem could be...?
Johnny
Johnny Billquist wrote:
On 2011-11-12 16.41, Oleg Safiullin wrote:
Huh. At what point? That should not happen. I don't see anything like
that when I try to log in as guest...
Just got the same error when i tried to log in as guest using SET HOST.
No problems for telnet login.
[...]
Oh! Darn! I bet I know what the problem is with that... The normal terminal driver only requires read access to the buffer that is used to write to the terminal. The RT: device driver, which is what is used for the CTERM, probably wants write access, even though it will only read from the buffer.
And then PIP (which is what TYPE invokes) does a QIO from a memory buffer that is read only... Not sure I can do anything to fix that... :-(
I'll test and check some more...
(I consider the write access check a bug in the device driver, and I've stumbled on it before when writing my own code...)
Johnny
Looks like even GCML doesn't work...
--
$ pip
PIP>
PIP -- Get Command Line - I/O Error
$ flx
FLX>
FLX -- I/O error on command input
$ run $mac
RT0>
RT0 -- Command I/O error
$
--
No problems on CTAKAH.
--
$ sh ter
RT0: [GUEST] [377,377] 12-NOV-11 22:50 1 A. GUEST
CLI = DCL BUF = 80. HFILL = 0
LINES = 24. TERM = VT2xx OWNER = RT0: NOPARITY
NOPRINTER_PORT NOPASTHRU NOSERIAL
LOWER NOPRIV NOHOLD NOSLAVE ESC CRT NOFORM REMOTE
NOECHO NOVFILL HHT NOFDX WRAP NORPA EBC NOTYPEAHEAD
NOCTRLC AVO ANSI DEC EDIT NOREGIS SOFT NOBLKMOD
HSYNC BRO NOABAUD TTSYNC
$ pip
PIP>^Z
$
--
On 2011-11-12 16.41, Oleg Safiullin wrote:
Huh. At what point? That should not happen. I don't see anything like
that when I try to log in as guest...
Just got the same error when i tried to log in as guest using SET HOST.
No problems for telnet login.
[...]
Oh! Darn! I bet I know what the problem is with that... The normal terminal driver only requires read access to the buffer that is used to write to the terminal. The RT: device driver, which is what is used for the CTERM, probably wants write access, even though it will only read from the buffer.
And then PIP (which is what TYPE invokes) does a QIO from a memory buffer that is read only... Not sure I can do anything to fix that... :-(
I'll test and check some more...
(I consider the write access check a bug in the device driver, and I've stumbled on it before when writing my own code...)
Johnny
--
...
$ @LB:[1,2]SYSLOGIN.CMD
What is your name? GHJ
RSX-11M-PLUS BL87mP (MIM ) 1292K PDP-11/74 at 16:37:59.5, 12-NOV-11
Users=7 Idle=100% Task=.IDLE. Up: 28 days, 21:56. Ticks=0 Errseq=15.
user task uic pri where state size(K) ioc par
GHJ AT.R0 * [ Self ] 64 RT0 Stopped 4/23 GEN
Johnny DTRT10 [ Self ] 50 TT10 Stopped 13/64 (1) GEN
J Billquist DTRT6 [ Oper ] 50 _TT6 Stopped 13/64 (1) GEN
GHJ WHOR0 * [ Self ] 50 RT0 Running 1/4 GEN
J Billquist (In MCR) [ Oper ] _TT4
J Billquist (In MCR) [ Oper ] _TT5
J Billquist (In MCR) [ Oper ] _TT7
J Billquist (In MCR) [ Oper ] _TT11
16:38:01 Task "TYPR0 " terminated on processor "A"
Memory protect violation
R0=000472
R1=027200
R2=000000
R3=000000
R4=001404
R5=000022
SP=000432
PC=147620
PS=070004
>
$
Short example, by the way:
$ dtr
DATATRIEVE-11, DEC Query and Report System
Version: V03.03-07, 1-DEC-89
Type HELP for help
DTR> ready hecnet
DTR> find hecnet with node_owner="Johnny Billquist"
[18 records found]
DTR> sort by area,node
DTR> print current
NODE NODE NODE REGISTER
NAME ADDRESS OWNER TIME
MAGICA 1.1 Johnny Billquist 11-Nov-2011
ERNIE 1.2 Johnny Billquist 11-Nov-2011
FNATTE 1.3 Johnny Billquist 11-Nov-2011
GOBLIN 1.4 Johnny Billquist 11-Nov-2011
ZEKE 1.5 Johnny Billquist 11-Nov-2011
GNOME 1.6 Johnny Billquist 11-Nov-2011
BJARNE 1.7 Johnny Billquist 11-Nov-2011
KRILLE 1.8 Johnny Billquist 11-Nov-2011
TEMPO 1.9 Johnny Billquist 11-Nov-2011
SIGGE 1.10 Johnny Billquist 11-Nov-2011
ROCKY 1.11 Johnny Billquist 11-Nov-2011
PONY 1.12 Johnny Billquist 11-Nov-2011
MIM 1.13 Johnny Billquist 11-Nov-2011
TINA 1.14 Johnny Billquist 11-Nov-2011
PONDUS 1.15 Johnny Billquist 11-Nov-2011
GNAT 1.16 Johnny Billquist 11-Nov-2011
JOSSE 1.17 Johnny Billquist 11-Nov-2011
BEA 1.19 Johnny Billquist 11-Nov-2011
DTR>
Johnny
On 2011-11-12 16.37, Johnny Billquist wrote:
On 2011-11-12 13.13, Rob Jarratt wrote:
I have tried logging in to MIM as guest, but I have never used either
Datatrieve or RSX and HELP does not show me anything that looks like it
might be Datatrieve. Can someone offer me some pointers to get started?
If you google for datatrieve, you'll find manuals. Mostly for VAX
Datatrieve, which is a superset, and also some manual for Datatrieve-11.
Those manuals are definitely more than enough to get you going. But you
can also just try starting it. The command is DTR, and there is a help
in there.
Thanks
Rob
NB When I logged in as Guest I got this:
13:04:53 Task "TYPR0 " terminated on processor "A"
Memory protect violation
R0=000472
R1=027200
R2=000000
R3=000000
R4=001404
R5=000022
SP=000432
PC=147620
PS=070004
Huh. At what point? That should not happen. I don't see anything like
that when I try to log in as guest...
Johnny
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE]
On Behalf Of hvlems at zonnet.nl
Sent: 11 November 2011 17:26
To: hecnet at Update.UU.SE
Subject: Re: [HECnet] Nodename database
Umm, write COBOL applications on top of DTR-11 and TDMS? :-) Hans ------
Origineel bericht------
Van: Johnny Billquist
Afzender: owner-hecnet at Update.UU.SE
Aan: hecnet at Update.UU.SE
Beantwoorden: hecnet at Update.UU.SE
Onderwerp: [HECnet] Nodename database
Verzonden: 11 november 2011 18:16
Hi. In my continuous effort of doing weird things, and finding new fun
ways of
playing around, I have now started setting up the nodename database
master
as a Datatrieve domain.
So, anyone who wants to peek at what I intend to use as master data can
connect to the Datatrieve server on MIM, and check out the HECNET
domain.
I
have not yet setup all the surrounding tools I plan to use to make
additions
and changes really easy to handle, but all current nodes should be in
that
storage anyway.
And if you are on MIM, you can of course also use Datatrieve locally.
Let me know if you have any questions or suggestions. I'm still
trying to
figure
out how to do this the neatest.
Johnny
Huh. At what point? That should not happen. I don't see anything like that when I try to log in as guest...
Just got the same error when i tried to log in as guest using SET HOST.
No problems for telnet login.
--
...
$ @LB:[1,2]SYSLOGIN.CMD
What is your name? GHJ
RSX-11M-PLUS BL87mP (MIM ) 1292K PDP-11/74 at 16:37:59.5, 12-NOV-11
Users=7 Idle=100% Task=.IDLE. Up: 28 days, 21:56. Ticks=0 Errseq=15.
user task uic pri where state size(K) ioc par
GHJ AT.R0 * [ Self ] 64 RT0 Stopped 4/23 GEN
Johnny DTRT10 [ Self ] 50 TT10 Stopped 13/64 (1) GEN
J Billquist DTRT6 [ Oper ] 50 _TT6 Stopped 13/64 (1) GEN
GHJ WHOR0 * [ Self ] 50 RT0 Running 1/4 GEN
J Billquist (In MCR) [ Oper ] _TT4
J Billquist (In MCR) [ Oper ] _TT5
J Billquist (In MCR) [ Oper ] _TT7
J Billquist (In MCR) [ Oper ] _TT11
16:38:01 Task "TYPR0 " terminated on processor "A"
Memory protect violation
R0=000472
R1=027200
R2=000000
R3=000000
R4=001404
R5=000022
SP=000432
PC=147620
PS=070004
>
$
On 2011-11-12 13.13, Rob Jarratt wrote:
I have tried logging in to MIM as guest, but I have never used either
Datatrieve or RSX and HELP does not show me anything that looks like it
might be Datatrieve. Can someone offer me some pointers to get started?
If you google for datatrieve, you'll find manuals. Mostly for VAX Datatrieve, which is a superset, and also some manual for Datatrieve-11. Those manuals are definitely more than enough to get you going. But you can also just try starting it. The command is DTR, and there is a help in there.
Thanks
Rob
NB When I logged in as Guest I got this:
13:04:53 Task "TYPR0 " terminated on processor "A"
Memory protect violation
R0=000472
R1=027200
R2=000000
R3=000000
R4=001404
R5=000022
SP=000432
PC=147620
PS=070004
Huh. At what point? That should not happen. I don't see anything like that when I try to log in as guest...
Johnny
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE]
On Behalf Of hvlems at zonnet.nl
Sent: 11 November 2011 17:26
To: hecnet at Update.UU.SE
Subject: Re: [HECnet] Nodename database
Umm, write COBOL applications on top of DTR-11 and TDMS? :-) Hans ------
Origineel bericht------
Van: Johnny Billquist
Afzender: owner-hecnet at Update.UU.SE
Aan: hecnet at Update.UU.SE
Beantwoorden: hecnet at Update.UU.SE
Onderwerp: [HECnet] Nodename database
Verzonden: 11 november 2011 18:16
Hi. In my continuous effort of doing weird things, and finding new fun
ways of
playing around, I have now started setting up the nodename database master
as a Datatrieve domain.
So, anyone who wants to peek at what I intend to use as master data can
connect to the Datatrieve server on MIM, and check out the HECNET domain.
I
have not yet setup all the surrounding tools I plan to use to make
additions
and changes really easy to handle, but all current nodes should be in that
storage anyway.
And if you are on MIM, you can of course also use Datatrieve locally.
Let me know if you have any questions or suggestions. I'm still trying to
figure
out how to do this the neatest.
Johnny
I have tried logging in to MIM as guest, but I have never used either
Datatrieve or RSX and HELP does not show me anything that looks like it
might be Datatrieve. Can someone offer me some pointers to get started?
Thanks
Rob
NB When I logged in as Guest I got this:
13:04:53 Task "TYPR0 " terminated on processor "A"
Memory protect violation
R0=000472
R1=027200
R2=000000
R3=000000
R4=001404
R5=000022
SP=000432
PC=147620
PS=070004
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE]
On Behalf Of hvlems at zonnet.nl
Sent: 11 November 2011 17:26
To: hecnet at Update.UU.SE
Subject: Re: [HECnet] Nodename database
Umm, write COBOL applications on top of DTR-11 and TDMS? :-) Hans ------
Origineel bericht------
Van: Johnny Billquist
Afzender: owner-hecnet at Update.UU.SE
Aan: hecnet at Update.UU.SE
Beantwoorden: hecnet at Update.UU.SE
Onderwerp: [HECnet] Nodename database
Verzonden: 11 november 2011 18:16
Hi. In my continuous effort of doing weird things, and finding new fun
ways of
playing around, I have now started setting up the nodename database master
as a Datatrieve domain.
So, anyone who wants to peek at what I intend to use as master data can
connect to the Datatrieve server on MIM, and check out the HECNET domain.
I
have not yet setup all the surrounding tools I plan to use to make
additions
and changes really easy to handle, but all current nodes should be in that
storage anyway.
And if you are on MIM, you can of course also use Datatrieve locally.
Let me know if you have any questions or suggestions. I'm still trying to
figure
out how to do this the neatest.
Johnny
Umm, write COBOL applications on top of DTR-11 and TDMS? :-)
Hans
------Origineel bericht------
Van: Johnny Billquist
Afzender: owner-hecnet at Update.UU.SE
Aan: hecnet at Update.UU.SE
Beantwoorden: hecnet at Update.UU.SE
Onderwerp: [HECnet] Nodename database
Verzonden: 11 november 2011 18:16
Hi. In my continuous effort of doing weird things, and finding new fun
ways of playing around, I have now started setting up the nodename
database master as a Datatrieve domain.
So, anyone who wants to peek at what I intend to use as master data can
connect to the Datatrieve server on MIM, and check out the HECNET
domain. I have not yet setup all the surrounding tools I plan to use to
make additions and changes really easy to handle, but all current nodes
should be in that storage anyway.
And if you are on MIM, you can of course also use Datatrieve locally.
Let me know if you have any questions or suggestions. I'm still trying
to figure out how to do this the neatest.
Johnny
Hi. In my continuous effort of doing weird things, and finding new fun ways of playing around, I have now started setting up the nodename database master as a Datatrieve domain.
So, anyone who wants to peek at what I intend to use as master data can connect to the Datatrieve server on MIM, and check out the HECNET domain. I have not yet setup all the surrounding tools I plan to use to make additions and changes really easy to handle, but all current nodes should be in that storage anyway.
And if you are on MIM, you can of course also use Datatrieve locally.
Let me know if you have any questions or suggestions. I'm still trying to figure out how to do this the neatest.
Johnny
I have succesfully copied VAX patches from SLAVE.
Now I wonder if this has something to do with the RHESUS problem:
Area State Circuit Next node to area
1 reachable ISA-0 1.300 (CTAKAH)
Should it be so?
Regards,
--Saku
On 2011-11-09 02.27, Saku Set l wrote:
I tried from GORVAX and it breaks the same way..
MAISA::SYSTEM$ dir gorvax::
Directory GORVAX::SYS$SPECIFIC:[FAL$SERVER]
MNENG1.SYS;1 MNENG2.SYS;1 NETSERVER.LOG;75
NETSERVER.LOG;74
NETSERVER.LOG;73 NETSERVER.LOG;72 NETSERVER.LOG;71
NETSERVER.LOG;70
PR0801ENG.SYS;1 PS0801ENG.SYS;1 SH1601ENG.SYS;1 WWENG1.SYS;1
Total of 12 files.
copy/log GORVAX::SYS$SPECIFIC:[FAL$SERVER]*.*.* []
%COPY-E-READERR, error reading GORVAX::SYS$SPECIFIC:[FAL$SERVER]MNENG1.SYS;1
-RMS-F-SYS, QIO system service request failed
-SYSTEM-F-LINKEXIT, network partner exited
%COPY-W-NOTCMPLT, GORVAX::SYS$SPECIFIC:[FAL$SERVER]MNENG1.SYS;1 not
completely copied
%COPY-E-CLOSEIN, error closing
GORVAX::SYS$SPECIFIC:[FAL$SERVER]MNENG1.SYS;1 as input
-RMS-F-WBE, error on write behind
-SYSTEM-F-LINKABORT, network partner aborted logical link
I just tried transferring a file from GORVAX to MIM, and it worked fine.
(Pulled SH1601ENG.SYS, just to test...)
Thinking a little more, I wonder if the problem could partially be that RHESUS might be running on gigabit ethernet, while GORVAX might be a 10 Mbit ethernet, unless it's a simulated VAX.
Thus, you can get into a state where packets are dropped along the way at several places. Both the local network, UDP, the bridge, the remote network, as well as the routers themself.
Sampsa could perhaps tell a little more about what the actual setup is at his end?
Johnny
Thanks Mark, I'll take a look.
--Saku
On Wed, Nov 9, 2011 at 1:53 PM, Mark Wickens <mark at wickensonline.co.uk> wrote:
Saku
SLAVE is now up.
You can browse the archive on the internet: http://hecnet.eu/archive
You can access via HECNET using the FAL$SERVER account, e.g: dir slave::[.patches]
Regards, Mark
On 09/11/11 11:24, Saku Set l wrote:
Hi Mark,
Actually I'm after the patches, but when I was looking for them I decided also to look at the VAX 7.3 layered products ISO image, since don't have that online right now..
--Saku
On Wed, Nov 9, 2011 at 1:12 PM, Mark Wickens <mark at wickensonline.co.uk>
What are you after Saku? It's probably on SLAVE, do you want to try and pull it from there just for comparison purposes?
Mark.
Saku
SLAVE is now up.
You can browse the archive on the internet: http://hecnet.eu/archive
You can access via HECNET using the FAL$SERVER account, e.g: dir slave::[.patches]
Regards, Mark
Hi Mark,
Actually I'm after the patches, but when I was looking for them I decided also to look at the VAX 7.3 layered products ISO image, since don't have that online right now..
--Saku
On Wed, Nov 9, 2011 at 1:12 PM, Mark Wickens <mark at wickensonline.co.uk>
What are you after Saku? It's probably on SLAVE, do you want to try and pull it from there just for comparison purposes?
Mark.
On 09/11/11 11:02, Saku Set l wrote:
I tried also from my Alpha with V8.3:
OpenVMS V8.3 on node DS20 9-NOV-2011 13:38:08.04 Uptime 266 18:41:29
No difference, same error.
--Saku
What are you after Saku? It's probably on SLAVE, do you want to try and pull it from there just for comparison purposes?
Mark.
Oh! I misunderstood. Ignore my comment...
I'll do some more experimentation instead.
Johnny
On 2011-11-09 11:41, Johnny Billquist wrote:
Hmm. GORVAX and RHESUS are on the same local ethernet segment, I would
think. That suggests that is indeed some problem with the I64 system?
Anyone have an Alpha system with VMS V8 who can try?
Johnny
On 2011-11-09 02:27, Saku Set l wrote:
I tried from GORVAX and it breaks the same way..
MAISA::SYSTEM$ dir gorvax::
Directory GORVAX::SYS$SPECIFIC:[FAL$SERVER]
MNENG1.SYS;1 MNENG2.SYS;1 NETSERVER.LOG;75
NETSERVER.LOG;74
NETSERVER.LOG;73 NETSERVER.LOG;72 NETSERVER.LOG;71
NETSERVER.LOG;70
PR0801ENG.SYS;1 PS0801ENG.SYS;1 SH1601ENG.SYS;1 WWENG1.SYS;1
Total of 12 files.
copy/log GORVAX::SYS$SPECIFIC:[FAL$SERVER]*.*.* []
%COPY-E-READERR, error reading
GORVAX::SYS$SPECIFIC:[FAL$SERVER]MNENG1.SYS;1
-RMS-F-SYS, QIO system service request failed
-SYSTEM-F-LINKEXIT, network partner exited
%COPY-W-NOTCMPLT, GORVAX::SYS$SPECIFIC:[FAL$SERVER]MNENG1.SYS;1 not
completely copied
%COPY-E-CLOSEIN, error closing
GORVAX::SYS$SPECIFIC:[FAL$SERVER]MNENG1.SYS;1 as input
-RMS-F-WBE, error on write behind
-SYSTEM-F-LINKABORT, network partner aborted logical link
--Saku
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
Hmm. GORVAX and RHESUS are on the same local ethernet segment, I would think. That suggests that is indeed some problem with the I64 system?
Anyone have an Alpha system with VMS V8 who can try?
Johnny
On 2011-11-09 02:27, Saku Set l wrote:
I tried from GORVAX and it breaks the same way..
MAISA::SYSTEM$ dir gorvax::
Directory GORVAX::SYS$SPECIFIC:[FAL$SERVER]
MNENG1.SYS;1 MNENG2.SYS;1 NETSERVER.LOG;75
NETSERVER.LOG;74
NETSERVER.LOG;73 NETSERVER.LOG;72 NETSERVER.LOG;71
NETSERVER.LOG;70
PR0801ENG.SYS;1 PS0801ENG.SYS;1 SH1601ENG.SYS;1 WWENG1.SYS;1
Total of 12 files.
copy/log GORVAX::SYS$SPECIFIC:[FAL$SERVER]*.*.* []
%COPY-E-READERR, error reading GORVAX::SYS$SPECIFIC:[FAL$SERVER]MNENG1.SYS;1
-RMS-F-SYS, QIO system service request failed
-SYSTEM-F-LINKEXIT, network partner exited
%COPY-W-NOTCMPLT, GORVAX::SYS$SPECIFIC:[FAL$SERVER]MNENG1.SYS;1 not
completely copied
%COPY-E-CLOSEIN, error closing
GORVAX::SYS$SPECIFIC:[FAL$SERVER]MNENG1.SYS;1 as input
-RMS-F-WBE, error on write behind
-SYSTEM-F-LINKABORT, network partner aborted logical link
--Saku
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
I tried from GORVAX and it breaks the same way..
MAISA::SYSTEM$ dir gorvax::
Directory GORVAX::SYS$SPECIFIC:[FAL$SERVER]
MNENG1.SYS;1 MNENG2.SYS;1 NETSERVER.LOG;75 NETSERVER.LOG;74
NETSERVER.LOG;73 NETSERVER.LOG;72 NETSERVER.LOG;71 NETSERVER.LOG;70
PR0801ENG.SYS;1 PS0801ENG.SYS;1 SH1601ENG.SYS;1 WWENG1.SYS;1
Total of 12 files.
copy/log GORVAX::SYS$SPECIFIC:[FAL$SERVER]*.*.* []
%COPY-E-READERR, error reading GORVAX::SYS$SPECIFIC:[FAL$SERVER]MNENG1.SYS;1
-RMS-F-SYS, QIO system service request failed
-SYSTEM-F-LINKEXIT, network partner exited
%COPY-W-NOTCMPLT, GORVAX::SYS$SPECIFIC:[FAL$SERVER]MNENG1.SYS;1 not completely copied
%COPY-E-CLOSEIN, error closing GORVAX::SYS$SPECIFIC:[FAL$SERVER]MNENG1.SYS;1 as input
-RMS-F-WBE, error on write behind
-SYSTEM-F-LINKABORT, network partner aborted logical link
--Saku
Did some additional checking. Doing a dir from MIM to MAISA actually do work. So there is something funny going on with RHESUS.
Looking with tcpdump at the UDP packets, the one thing that looks a little different is that MAISA actually sends 584 byte length packets. Thinking about it, 609 is a rather odd value.
Since RHESUS is an I64, I'm ready to blame that modern junk. :-)
I can't spot what the problem is right now anyway. More to follow, I guess. Anyone else feel like trying to figure it out?
Johnny
On 2011-11-08 21.25, Johnny Billquist wrote:
Tried from MIM, and I can't even do a directory.
Got a "remote disconnect or line failure" error.
But I can talk NICE to RHESUS... Looking at MIM counters, everything
looks fine. Looking at RHESUS counters, it has plenty of response timeouts.
So it appears the data from RHESUS to MIM gets lost sometimes. In this
case just for doing a directory on RHESUS...
Looking at the UDP traffic, it looks like the packets are coming in to
Update, so it might be an issue with E11 (I've had an issue before where
it could drop ethernet packets under some circumstances).
But I'm not sure why it fails for you, Saku, since your traffic should
not pass by MIM anyway, even though it do pass through Update.
But anyway, it (for me) definitely does not look like I have problems
with packets being thrown away. Actually, the packet size of individual
UDP packets is 609 bytes. DECnet normally do not send lager packets than
defined by the EXEC parameter "segment buffer size" (or "buffer size" in
VMS). And the default for this is 576 bytes. Add some overhead to that,
and 609 bytes sounds reasonable.
Johnny
On 2011-11-08 17.50, Saku Set l wrote:
Hi,
First I tried ISO images from SYS$SPECIFIC:[FAL$SERVER.MEDIALIB]
but since it didn't work I tried smaller files from [FAL$SERVER] but it
still didn't work out. Directory comes out fine.
Regards,
--Saku
On Tue, Nov 8, 2011 at 3:06 PM, Johnny Billquist <bqt at softjar.se
<mailto:bqt at softjar.se>> wrote:
On 2011-11-08 05.38, Saku Set l wrote:
Sampsa&Co,
I tried to pull some files from RHESUS media library, but don't get
anything, just
-SYSTEM-F-LINKABORT, network partner aborted logical link
I have no problem on copying files from MIM.
MTU issues on the internet uplink?
Nope. Whatever it is, I seriously doubt that is it.
Unless some firewall blocks fragmenting, the MTU is not really ever
an issue.
The error might be because of excessive lost packets. Maybe some
throttling could help. What files are you trying to pull? I can try
and experiment some.
Johnny
Tried from MIM, and I can't even do a directory.
Got a "remote disconnect or line failure" error.
But I can talk NICE to RHESUS... Looking at MIM counters, everything looks fine. Looking at RHESUS counters, it has plenty of response timeouts.
So it appears the data from RHESUS to MIM gets lost sometimes. In this case just for doing a directory on RHESUS...
Looking at the UDP traffic, it looks like the packets are coming in to Update, so it might be an issue with E11 (I've had an issue before where it could drop ethernet packets under some circumstances).
But I'm not sure why it fails for you, Saku, since your traffic should not pass by MIM anyway, even though it do pass through Update.
But anyway, it (for me) definitely does not look like I have problems with packets being thrown away. Actually, the packet size of individual UDP packets is 609 bytes. DECnet normally do not send lager packets than defined by the EXEC parameter "segment buffer size" (or "buffer size" in VMS). And the default for this is 576 bytes. Add some overhead to that, and 609 bytes sounds reasonable.
Johnny
On 2011-11-08 17.50, Saku Set l wrote:
Hi,
First I tried ISO images from SYS$SPECIFIC:[FAL$SERVER.MEDIALIB]
but since it didn't work I tried smaller files from [FAL$SERVER] but it
still didn't work out. Directory comes out fine.
Regards,
--Saku
On Tue, Nov 8, 2011 at 3:06 PM, Johnny Billquist <bqt at softjar.se
<mailto:bqt at softjar.se>> wrote:
On 2011-11-08 05.38, Saku Set l wrote:
Sampsa&Co,
I tried to pull some files from RHESUS media library, but don't get
anything, just
-SYSTEM-F-LINKABORT, network partner aborted logical link
I have no problem on copying files from MIM.
MTU issues on the internet uplink?
Nope. Whatever it is, I seriously doubt that is it.
Unless some firewall blocks fragmenting, the MTU is not really ever
an issue.
The error might be because of excessive lost packets. Maybe some
throttling could help. What files are you trying to pull? I can try
and experiment some.
Johnny
Hi,
First I tried ISO images from SYS$SPECIFIC:[FAL$SERVER.MEDIALIB]
but since it didn't work I tried smaller files from [FAL$SERVER] but it still didn't work out. Directory comes out fine.
Regards,
--Saku
On Tue, Nov 8, 2011 at 3:06 PM, Johnny Billquist <bqt at softjar.se> wrote:
On 2011-11-08 05.38, Saku Set l wrote:
Sampsa&Co,
I tried to pull some files from RHESUS media library, but don't get
anything, just
-SYSTEM-F-LINKABORT, network partner aborted logical link
I have no problem on copying files from MIM.
MTU issues on the internet uplink?
Nope. Whatever it is, I seriously doubt that is it.
Unless some firewall blocks fragmenting, the MTU is not really ever an issue.
The error might be because of excessive lost packets. Maybe some throttling could help. What files are you trying to pull? I can try and experiment some.
Johnny
On 2011-11-08 05.38, Saku Set l wrote:
Sampsa&Co,
I tried to pull some files from RHESUS media library, but don't get
anything, just
-SYSTEM-F-LINKABORT, network partner aborted logical link
I have no problem on copying files from MIM.
MTU issues on the internet uplink?
Nope. Whatever it is, I seriously doubt that is it.
Unless some firewall blocks fragmenting, the MTU is not really ever an issue.
The error might be because of excessive lost packets. Maybe some throttling could help. What files are you trying to pull? I can try and experiment some.
Johnny
