SERIOUS ERROR LDAP Daemon (LDAP Listener )

[dennis]

I posted this in another thread however maybe should have posted it here.

Any thoughts on a fix for this. After it crashes no logins allowed. The most recent time required me to reboot the server.

We are running scalix-server version 11.0.0.431 release 1.sles9

SERIOUS ERROR LDAP Daemon (LDAP Listener ) 01.27.07 02:58:51
[OM 10270] Process about to terminate due to error.
Signal (Segmentation Violation) trapped by process 21741
Procedure trace follows:

SERIOUS ERROR LDAP Daemon (LDAP Listener ) 01.27.07 02:58:51
[OM 10272] BACKTRACE:
/opt/scalix/lib/libom_er.so(er_add_backtrace+0xc6)[0x4001c366]
/opt/scalix/lib/libom_er.so[0x4001c665]
/opt/scalix/lib/libom_er.so(er_DumpProcAndExit+0x1f)[0x4001c7ef]
[0xffffe420]
/lib/tls/libpthread.so.0[0x400b8cf7]
/lib/tls/libc.so.6(__clone+0x5e)[0x4017e21e]

[florian]

1. does this happen in connection with any user activity? (the most likely would be address book lookups from SWA, because these go through LDAP)

2. if this happens, can you restart the LDAP server by typing "omon -w ldap"

Thx,
Florian.

[dennis]

This is a very new installation so i do not believe address book lookups are the issue although i'm not completely sure.

Only 2 or three users were in the webmail and of those 3, 2 were admin. The other was a user i was on the phone with.

This was the second time it failed. I was able to click start the first time it happend in the SWA however i was already logged into it.

The second time i was unable to log into the SWA to start it. However i did not try the command line start. SSH was not routed in at the time so i rebooted using webmin.

[jch]

It's a pity you didn't include the prodedure trace from the log message -- was that because it wasn't there or did you deliberately exclude it?

Are you fully up to date with patches from SuSE? (yast2 online_update) Sometimes it's just a problem in some underlying system library that's been fixed at some stage, but if you're on the latest everything then at least we know where we stand.

Possibly a way to try to find out what's going on it to stop the LDAP server (omoff -d0 -w slapd) and then run it in debug mode (omslapd -d13).

This produces lots of output, but we're really only interested in what was produced shortly before it crashed. You can see the individual LDAP queries being processed and you should be able to tell what was going on when it crashed.

jch

[dennis]

Sorry, i only posted what was listed in the eventlog in the SWA. I'll get the log you refer to and post it plus confirm the updates with SUSE.

[florian]

LDAP lookups are also connected to activity within SAC, so you won't be able to use SAC when LDAP is down.

Just wondering - do you know if any other LDAP related software is running on this system, e.g.. an open ldap server or is the systems OS login and authentication setup to work against an external LDAP server?

Cheers,
Florian.

[dennis]

No, the only ldap process is the only included with Scalix.
5022 ? Ss 0:15 /opt/scalix/bin/ldapmapper --pid-file /var/run/ldapmapper.pid


There are are some SuSE updates as well which I will install now:

"Security update for XFree8x
"Security update for IBM Java2 JRE and SDK
"Recommended update for udev
"Security update for gtk2
"Recommended update for hwinfo
"Security update for bind

Just to be clear, which log file are you referring to? We are pulling the logfile from the SAC.

[florian]

it's probably more efficient to pull logfile information from the command line...

you might want to check

Code: Select all

cat /var/opt/scalix/??/s/logs/fatal

Code: Select all

cat /var/opt/scalix/??/s/logs/daemon.stderr

and

Code: Select all

omshowlog -l 7 -s ldap -p 30

(the number after the p gives you messages from the last 30 minutes only, so should be adjusted to how long after the crash you execute the command)

These should give you logs with full stack traces.

Florian.

[dennis]

I'll submit the log in a PM to you based on it contains username and some other errors maybe we can work though.

[dennis]

Florian, i just posted a pm with a link to grab the zip file containing the requested logs.

Best Regards,
Dennis

[jch]

Florian passed the logs on to me -- thanks.

The logs don't really tell us anything other than the fact you might be having problems with your DNS -- I wouldn't rule it out, but I don't suppose that's the problem you're seeing.

You need to run the LDAP daemon in debug mode and look at what it produces just before it crashes so we can get a handle on what it was doing just before it crashed.

jch

[dennis]

jch - thank you.

i didnt think we were having a dns issue? would you please let me know when you have a chance what the error was?

I'll get the LDAP Debug going as well. As soon as i have it i'll update this thread and send you a pm with the link to the logs.

Question: i have another thread going regarding the imap issue we had. Would you like these logs as well? viewtopic.php?t=5939

[jch]

One of the log files showed a lot of name resolution issues, I forget which one, but it was pretty obvious.

I don't want the complete debug output, just the last few requests' worth so we can see what the server was doing when it crashed.

jch