Wonderful x25

[Paula J Frazer-Campbell]

Hi to all.

My problem is now fixed by a rebuild of the x25/x29/x3 software, but I am now in the phase of tracing what when wrong.

FAULT.

All x25 users were stopping.

x29 server sessions were building up but no through connectivity.

Logging in as x25 users - /etc/issue info displayed and then login stops.

The users were not getting to the application section of x29hosts ????

Question :- What happens between the /etc/issue screen on a x25 user and the x29hosts file ???

Any Ideas gratfully received.

Paula

[David Navarro]

Maybe something with x3config file

Take a look at this site:

http://www.informatik.uni-frankfurt.de/doc/man/hpux/x3config.4.html

[Paula J Frazer-Campbell]

Thanks David

But all my files hav not changed for month and I take approx 14000 connections each day on these connections.

What I am trying to figue is why it stopped where it did as on rebuilding the connection software I used sam and just entered my documented settings and it worked ???

>:-(

Paula


Paula

[David Navarro]

Are you sure you're using these files?
Years ago, I was installing a X25 connection, and I do everything manually, and when I try to do something with sam, it stop to work, I remember the problem was something like I was using a file named x25_config, and sam change it for use another named x25_config_0 or similar.

[Anthony deRito · Accepted solution]

Paula, there are conditions which must be met before a service request will be accepted by the x29server. They are listed in the man page for x29server.

Also assuming that your link was OK, can it be that your application listing in pad_spt{} was not responding due to some resource issue? (You probably have already ruled this out.)

When you say you ?rebuild the connection software? do you mean the you re-declared the X.25 interface parameters with SAM and then restarted the interface? If this is the case, then a lot happened during the time the interface was re-initialized. You basically went through and set up layer 1, 2 and 3 protocols for X.25 (PLP, LAPB). A lot happened here and it is hard to determine if something broke during the call. The trick would be to rule the problem from call setup mode all the way to call clearing mode. The last operational mode of X.25 is called restarting. This is the mode that probably fixed your problem.

This would be done by verifying the connections at all 3 levels that X.25 goes through. X.25 includes protocols at level 1, 2 and 3. (OSI reference model). One way would have been to run:

X25ifstate ?d [dev file]

This would have told you if the X.25 link was disconnected at level 1, 2 or 3.

Another way to do this would be to use the x25check command. X25check would have allowed you to interactively check communication at all 3 levels. I?ve never used this one but it looks helpful.

Since your users did get to /etc/issue, you would think that level 3 was OK but there may have still been a breakdown down lower.

Lastly, you can also set up nettl to capture X.25 stats in case it happens again.

Hope this helps.

Tony

[Paula J Frazer-Campbell]

Tony,

Thanks for the info, very helpful.

When I was checking the x25 connectivity I contacted my service supplier and we went through a testing phase in which they could see viable data packets from us to them?

The error that emerged was a clear from the DT End - from my server.

With the x25 in its problem state I could connect as a normal user - the /etc/issue was displayed and then the connection stopped - by pressing return I could then connect all the way to our application?

As you surmised I had already checked the application and had executed the application line from x29hosts ? with no errors.

I am therefore trying to figure at what stage /etc/issue is called in the x25 initialisation and what occurs afterwards that a return would clear?

>>:-((

Paula