ucx telnet access, i'm missing something....

[cgilley]

This has got to fall into the forehead slapping zone - I'm moving a VAX (5.5, running ucx) to a new network area.  But the users cannot telnet in to the vax using putty.  This is a very mature system, and I'm sure I have forgotten the obvious.  Bot of these "vaxes" are running in VMs using charon-vax.

From my laptop, I can ping both servers hosting the VMs, and I can ping the existing VAX, but the new vax remains unpingable.  This leads to my inability to telnet to it.

Something deep inside of me says their is something I forgot to do in ucx ...

what am I missing?

appreciate the obvious

[Steven Schweda]

> [...] VAX (5.5, running ucx) [...]

      ucx show version

> [...] to a new network area. [...]

   Define "new".  What are the IP addresses of these systems?

> [...] users cannot telnet in to the vax using putty. [...]

   "cannot" is not a useful problem description.  An actual error
message might be more helpful.

> what am I missing?

   Can't tell without some info on the network configuration.  The VMS
system could simply have an IP address in the wrong subnet, and no route
to the right subnet.

[cgilley]

Not a mind reader?  i'm disappointed.  lol, yeah, I didn't provide much information. 

This is an old legacy system running OpenVMS 6.1, UCX 4.0 within the charon-vax emulation environment.  I'm in the process of migrating to a new virtual machine running in the customer's cloud.  The new installation is using the exact same disk files as the existing system.

The existing system is at 10.90.10.202.  I can access the system from my laptop.

I took the existing disk files, moved them to the VM and simply used UCX$CONFIG to change the new system's ip address to 10.90.10.186.  Can't ping this address, can't telnet into the machine.

The addresses are so close, I would not expect routing to pay an issue.

Forehead slap moment - Windows Server 2012 is probably doing this, not the vax.... I'll be back to close this issue out.

[Volker Halle]

You need to also look at the CHARON-VAX network configuration and VMware network configuration !

Is the CHARON-VAX network interface correctly mapped to the dedicated Windows network adapter to be used by CHARON exclusively ? NDIS6 binding o.k. ?

Looking from VMS: do the counters of $ UCX SHOW INTERFACE show packets being sent and/or received ?

Does the CHARON NCC tool show packets being sent and/or received ?

Does the VAX use DECnet on the LAN interface (in addition to IP) ? If so, is the VMware LAN interface correctly configured (MAC address change enabled and promiscuous mode enabled) ?

Volker.

[cgilley]

Yes to all.  And good suggestions, thanks. The part I missed is that if I connect to the Server 2012 machine, I have no problem telneting or pinging the VAX.  I think I'm getting hammered by a group policy out in the customer's domain.

[Chris Scheers]

IIRC, Server 2012 enables ECN by default. ECN can interfere with some older TCP/IP stacks as its definition is not strictly upwards compatible. You might see if disabling ECN makes any difference. Also, do your testing on a box other than the host box so that the bits have to actually travel over a wire. If you are trying to communicate from the host box to an emulated/VM machine, TOE issues may come into play.

[Brad McCusker]

I agree with prior statements, this sounds like a network problem on the host.  There are lots of considerations for networking CHARON in the cloud.  Follow the Stromasys documentation closely.  

[cgilley]

Turns out it's death by typos.  I changed the IP address and missed the route entry.

The route entry was left to the old server IP address....

thanks for the suggestions