[Simh] VAX DZ terminal controller unit?

Bucher, Andreas (Andreas)** CTR ** andreas.bucher at alcatel-lucent.com
Fri May 20 13:43:58 EDT 2011


Hi Gary, 

this reminds me to something :-)

It might not just be an "issue" with SIMH, but with any "real" vax as well. I remember spending lots of time debugging terminal lines on a 11/750, because it gave me the same behaviour :-) Only I don't quite remeber what I did to cure it.


I just tried it on SIMH (V3.7-1, VMS 7.2) - similar configuration in VAX.INI (16 lines as default, attach DZ 1031): exactly as you describe.

I am still searchin' my memory for the trick to cure the behaviour :-) 

The only recommendation I could come up with so far is to kill the login process that waits for the terminal line to become active. This seems to get stuck at every first attempt after a logoff. Maybe it somehow really misses the disconnect of the terminal.

So: 
Assuming you are logged in to a Telnet session connected to TTA1:

- log off that session
- hit ENTER (nothing will happen, but the VAX will fire up another loginout.exe image attached to that terminal port and device)
- on another session, perhaps the console, issue
$ stop _tta1:
to kill that process.

- Now back on the TTA1 telnet session, it should be possible to hit ENTER again and get the login prompt immediately.

I just tried it several times, it works all the time, and frees up the terminal port immediately. The disadvantage is that this has to be done each time you want to re-use a terminal (port). 



As a workaround, you could set up an automatic job that monitors the owner status of a terminal port.
Use
$ show terminal tta1
to determine the status. If it says "Owner: No owner", nobody is logged on and no terminal is connected, thus, no loginout.exe is running on that port.

If you hit ENTER once on TTA1 at that point, it changes to "Owner: _TTA1:", which is the process name. 
Below, it states "Username: <login>. If the scheduled job now kills the owner process if that state persists for some time, it will free up this terminal automatically. Looping thru all terminals would clean up everything.


Nevertheless, this is a crude workaround only, there should be either some settings in VMS to change that behaviour, or something in SIMH is not behaving correctly, hiding the disconnected state of the terminal and making VMS thing that nothing changed.

Regards,
Andreas



> -----Original Message-----
> From: simh-bounces at trailing-edge.com 
> [mailto:simh-bounces at trailing-edge.com] On Behalf Of Gary Lee Phillips
> Sent: Thursday, May 19, 2011 6:59 PM
> To: simh at trailing-edge.com
> Subject: [Simh] VAX DZ terminal controller unit?
> 
> I'm running SIMH 3.8 on Ubuntu 10.04 with OpenVMS 7.3 from the OpenVMS
> Hobbyist CD.
> 
> The emulation seems to be working well for almost everything I've
> tried, except that using the DZ device causes some odd behavior.
> 
> In my vax.ini file I have:
> 
> SET DZ LINES=8
> ATTACH DZ 8501
> 
> When the VAX emulation is booted, it is possible to "telnet localhost
> 8501" (or use a telnet client from another machine to port 8501 on the
> emulation host) and get a login prompt. This works as expected for one
> session. However, after logging out it is no longer possible to
> connect again. When I try, I get as far as
> 
> Connected to the VAX simulator DZ device, line 0
> 
> But pressing ENTER or CTRL-P or any other key never brings up a login
> prompt. Sometimes after a few minutes it will work again, but not
> always. If I leave the "dead" session connected and try to start
> another instance of the telnet client, that connects to "line 1" and
> works, but once it is logged out, line 1 is no longer usable. And so
> forth...
> 
> Something is not being reinitialized. I've tried "logout/hangup" and
> "logout/nohangup" and using the telnet client to send break signals
> and other out-of-band communications, but nothing seems to revive the
> line.
> 
> I may well be missing something here. Any advice?
> 
> Note that with TCP/IP installed and working on the OpenVMS system, a
> standard telnet connection to the emulator still works correctly, but
> this doesn't allow a terminal connection from the emulator host
> machine itself. It only works from another machine since both IP
> addresses are hosted on a single NIC in the emulator host.
> 
> Thanks for any help you might offer.
> 
> Gary
> _______________________________________________
> Simh mailing list
> Simh at trailing-edge.com
> http://mailman.trailing-edge.com/mailman/listinfo/simh
> 


More information about the Simh mailing list