On 2012-12-28 22:42, Cory Smelosky wrote:
On 28 Dec 2012, at 16:35, Cory Smelosky <b4 at gewt.net> wrote:
On 28 Dec 2012, at 16:33, "Bob Armstrong" <bob at jfcl.com> wrote:
Johnny wrote:
VMS don't know about bridging, ...
Are we talking about DECnet here? If so, then that's not true. VMS is
perfectly happy with two NI interfaces. Both have the same DECnet address
and therefore the same MAC and are presumably connected to different
physical networks. It'd be a mess if you connected them both to the same
LAN segment, of course. IF the VMS node is a router, it'll happily route
DECnet traffic between the two interfaces and their associated networks.
I'm doing it now - LEGATO is configured this way.
I'm starting to think the problem lies with simh interfacing with FreeBSD's netstack/some odd OpenVPN stuff.
I fixed it. Turns out the problem resulted from accidentally becoming the router for someone else's network.
(Feel free to ask how the hell I managed that)
Please tell. :-)
Johnny
On 2012-12-28 22:33, Bob Armstrong wrote:
Johnny wrote:
VMS don't know about bridging, ...
Are we talking about DECnet here? If so, then that's not true. VMS is
perfectly happy with two NI interfaces. Both have the same DECnet address
and therefore the same MAC and are presumably connected to different
physical networks. It'd be a mess if you connected them both to the same
LAN segment, of course. IF the VMS node is a router, it'll happily route
DECnet traffic between the two interfaces and their associated networks.
I'm doing it now - LEGATO is configured this way.
And I'd like to point out that what you describe is not briding... :-)
Yes, VMS can route packets. Yes, all ethernet interfaces under DECnet phase IV will have the same MAC address. No, VMS have no idea of the concept of a bridge. And if you bridge two ethernet segments together, and VMS sits on both those segments, bad juju happens.
Johnny
Johnny Billquist <bqt at softjar.se> writes:
On 2012-12-28 21:00, Cory Smelosky wrote:
Hello!
%NCP-W-OPEFAI, Operation failure
%SYSTEM-F-IVADDR, invalid media address
%NCP-W-UNRCMP, Unrecognized component , Circuit
Circuit = QNA-1
However:
XQ, address=20001920-2000192F*, vector=250, MAC=AA:00:04:00:15:24, type=DELQA-T, poll=100, attached to em0
===================================================^^^^^^^^^^^^^^^^^
XQB, address=20001930-2000193F*, vector=2E0, MAC=AA:00:04:00:15:24, type=DELQA-T, poll=100, attached to tap0
====================================================^^^^^^^^^^^^^^^^^
DECnet modifies the ethernet address of the interface. If these sit
on a common ethernet, you're going to have problems! Perhaps, that's
what you are seeing. Trying setting the second circuit to a unique
address. Try address 9.22 on the second ethernet circuit.
--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG
Well I speak to machines with the voice of humanity.
On 28 Dec 2012, at 16:35, Cory Smelosky <b4 at gewt.net> wrote:
On 28 Dec 2012, at 16:33, "Bob Armstrong" <bob at jfcl.com> wrote:
Johnny wrote:
VMS don't know about bridging, ...
Are we talking about DECnet here? If so, then that's not true. VMS is
perfectly happy with two NI interfaces. Both have the same DECnet address
and therefore the same MAC and are presumably connected to different
physical networks. It'd be a mess if you connected them both to the same
LAN segment, of course. IF the VMS node is a router, it'll happily route
DECnet traffic between the two interfaces and their associated networks.
I'm doing it now - LEGATO is configured this way.
I'm starting to think the problem lies with simh interfacing with FreeBSD's netstack/some odd OpenVPN stuff.
I fixed it. Turns out the problem resulted from accidentally becoming the router for someone else's network.
(Feel free to ask how the hell I managed that)
Bob
--
Cory Smelosky
http://gewt.net/ Personal stuff!
http://gimme-sympathy.org/ My permanently-a-work-in-progress pet project.
--
Cory Smelosky
http://gewt.net/ Personal stuff!
http://gimme-sympathy.org/ My permanently-a-work-in-progress pet project.
On 28 Dec 2012, at 16:33, "Bob Armstrong" <bob at jfcl.com> wrote:
Johnny wrote:
VMS don't know about bridging, ...
Are we talking about DECnet here? If so, then that's not true. VMS is
perfectly happy with two NI interfaces. Both have the same DECnet address
and therefore the same MAC and are presumably connected to different
physical networks. It'd be a mess if you connected them both to the same
LAN segment, of course. IF the VMS node is a router, it'll happily route
DECnet traffic between the two interfaces and their associated networks.
I'm doing it now - LEGATO is configured this way.
I'm starting to think the problem lies with simh interfacing with FreeBSD's netstack/some odd OpenVPN stuff.
Bob
--
Cory Smelosky
http://gewt.net/ Personal stuff!
http://gimme-sympathy.org/ My permanently-a-work-in-progress pet project.
Johnny wrote:
VMS don't know about bridging, ...
Are we talking about DECnet here? If so, then that's not true. VMS is
perfectly happy with two NI interfaces. Both have the same DECnet address
and therefore the same MAC and are presumably connected to different
physical networks. It'd be a mess if you connected them both to the same
LAN segment, of course. IF the VMS node is a router, it'll happily route
DECnet traffic between the two interfaces and their associated networks.
I'm doing it now - LEGATO is configured this way.
Bob
On 2012-12-28 22:04, Cory Smelosky wrote:
On 28 Dec 2012, at 16:03, Johnny Billquist <bqt at softjar.se> wrote:
I didn't ask what host you are running your simulator on, I asked what OS you are running the thing that talks DECnet.
But I now suspect you are playing with VMS. Afraid I can't help much with that one. In theory there shouldn't be any problems, but exactly why are getting the errors I can only guess.
And my first guess would be that the CSR is wrong.
My second guess would be that something is broken in simh, or VMS, if I look at the error message, since it seems to complain about the MAC address. VMS should be fine with this, however, if it does some verification of the MAC address before trying to change it, then you might have a problem. (Naturally, no two interfaces should ever have the same default MAC address.)
Actually they might due to the bridging. I'll go generate a random MAC and see if that helps any.
VMS don't know about bridging, so from VMS view of the world, no two interfaces should ever have the same default address.
Note that VMS will change the address as a part of the initialization, and of course you will have the same address on the two interfaces after that.
The problem is that simh might not properly handle a changed MAC address on an interface, which is why you are recommended to set it to the "right" value already before booting VMS.
Anyway, I'm really making lots of guesses here right now, so don't take what I write as any absolute truths. Good luck.
Johnny
Johnny
On 2012-12-28 21:57, Cory Smelosky wrote:
On 28 Dec 2012, at 15:56, Johnny Billquist <bqt at softjar.se> wrote:
On 2012-12-28 21:00, Cory Smelosky wrote:
Hello!
%NCP-W-OPEFAI, Operation failure
%SYSTEM-F-IVADDR, invalid media address
%NCP-W-UNRCMP, Unrecognized component , Circuit
Circuit = QNA-1
However:
XQ, address=20001920-2000192F*, vector=250, MAC=AA:00:04:00:15:24, type=DELQA-T, poll=100, attached to em0
XQB, address=20001930-2000193F*, vector=2E0, MAC=AA:00:04:00:15:24, type=DELQA-T, poll=100, attached to tap0
QNA-0 comes up fine.
Anything special I need to do? I NEED to be using two NICs and two circuits for this.
It probably helps if you tell which OS, and what kind of DECnet you have configured. :-)
FreeBSD. I'm trying to route DECnet packets coming on XQB to a Multinet tunnel established via XQA. It needs to route area 9.
I /might/ have just found a workaround that's a little bit crazy. ;)
Johnny
--
Cory Smelosky
http://gewt.net/ Personal stuff!
http://gimme-sympathy.org/ My permanently-a-work-in-progress pet project.
--
Cory Smelosky
http://gewt.net/ Personal stuff!
http://gimme-sympathy.org/ My permanently-a-work-in-progress pet project.
On 28 Dec 2012, at 16:03, Johnny Billquist <bqt at softjar.se> wrote:
I didn't ask what host you are running your simulator on, I asked what OS you are running the thing that talks DECnet.
But I now suspect you are playing with VMS. Afraid I can't help much with that one. In theory there shouldn't be any problems, but exactly why are getting the errors I can only guess.
And my first guess would be that the CSR is wrong.
My second guess would be that something is broken in simh, or VMS, if I look at the error message, since it seems to complain about the MAC address. VMS should be fine with this, however, if it does some verification of the MAC address before trying to change it, then you might have a problem. (Naturally, no two interfaces should ever have the same default MAC address.)
Actually they might due to the bridging. I'll go generate a random MAC and see if that helps any.
Johnny
On 2012-12-28 21:57, Cory Smelosky wrote:
On 28 Dec 2012, at 15:56, Johnny Billquist <bqt at softjar.se> wrote:
On 2012-12-28 21:00, Cory Smelosky wrote:
Hello!
%NCP-W-OPEFAI, Operation failure
%SYSTEM-F-IVADDR, invalid media address
%NCP-W-UNRCMP, Unrecognized component , Circuit
Circuit = QNA-1
However:
XQ, address=20001920-2000192F*, vector=250, MAC=AA:00:04:00:15:24, type=DELQA-T, poll=100, attached to em0
XQB, address=20001930-2000193F*, vector=2E0, MAC=AA:00:04:00:15:24, type=DELQA-T, poll=100, attached to tap0
QNA-0 comes up fine.
Anything special I need to do? I NEED to be using two NICs and two circuits for this.
It probably helps if you tell which OS, and what kind of DECnet you have configured. :-)
FreeBSD. I'm trying to route DECnet packets coming on XQB to a Multinet tunnel established via XQA. It needs to route area 9.
I /might/ have just found a workaround that's a little bit crazy. ;)
Johnny
--
Cory Smelosky
http://gewt.net/ Personal stuff!
http://gimme-sympathy.org/ My permanently-a-work-in-progress pet project.
--
Cory Smelosky
http://gewt.net/ Personal stuff!
http://gimme-sympathy.org/ My permanently-a-work-in-progress pet project.
I didn't ask what host you are running your simulator on, I asked what OS you are running the thing that talks DECnet.
But I now suspect you are playing with VMS. Afraid I can't help much with that one. In theory there shouldn't be any problems, but exactly why are getting the errors I can only guess.
And my first guess would be that the CSR is wrong.
My second guess would be that something is broken in simh, or VMS, if I look at the error message, since it seems to complain about the MAC address. VMS should be fine with this, however, if it does some verification of the MAC address before trying to change it, then you might have a problem. (Naturally, no two interfaces should ever have the same default MAC address.)
Johnny
On 2012-12-28 21:57, Cory Smelosky wrote:
On 28 Dec 2012, at 15:56, Johnny Billquist <bqt at softjar.se> wrote:
On 2012-12-28 21:00, Cory Smelosky wrote:
Hello!
%NCP-W-OPEFAI, Operation failure
%SYSTEM-F-IVADDR, invalid media address
%NCP-W-UNRCMP, Unrecognized component , Circuit
Circuit = QNA-1
However:
XQ, address=20001920-2000192F*, vector=250, MAC=AA:00:04:00:15:24, type=DELQA-T, poll=100, attached to em0
XQB, address=20001930-2000193F*, vector=2E0, MAC=AA:00:04:00:15:24, type=DELQA-T, poll=100, attached to tap0
QNA-0 comes up fine.
Anything special I need to do? I NEED to be using two NICs and two circuits for this.
It probably helps if you tell which OS, and what kind of DECnet you have configured. :-)
FreeBSD. I'm trying to route DECnet packets coming on XQB to a Multinet tunnel established via XQA. It needs to route area 9.
I /might/ have just found a workaround that's a little bit crazy. ;)
Johnny
--
Cory Smelosky
http://gewt.net/ Personal stuff!
http://gimme-sympathy.org/ My permanently-a-work-in-progress pet project.
On 28 Dec 2012, at 15:56, Johnny Billquist <bqt at softjar.se> wrote:
On 2012-12-28 21:00, Cory Smelosky wrote:
Hello!
%NCP-W-OPEFAI, Operation failure
%SYSTEM-F-IVADDR, invalid media address
%NCP-W-UNRCMP, Unrecognized component , Circuit
Circuit = QNA-1
However:
XQ, address=20001920-2000192F*, vector=250, MAC=AA:00:04:00:15:24, type=DELQA-T, poll=100, attached to em0
XQB, address=20001930-2000193F*, vector=2E0, MAC=AA:00:04:00:15:24, type=DELQA-T, poll=100, attached to tap0
QNA-0 comes up fine.
Anything special I need to do? I NEED to be using two NICs and two circuits for this.
It probably helps if you tell which OS, and what kind of DECnet you have configured. :-)
FreeBSD. I'm trying to route DECnet packets coming on XQB to a Multinet tunnel established via XQA. It needs to route area 9.
I /might/ have just found a workaround that's a little bit crazy. ;)
Johnny
--
Cory Smelosky
http://gewt.net/ Personal stuff!
http://gimme-sympathy.org/ My permanently-a-work-in-progress pet project.
