Remote Client Interface Died

[dougp23]

Apparently.

Unable to login via webmail with: (paraphrasing here) currently unavailable. try again later.

Couldn't login to sac, it just sat there with Attempting to Login.

After a morning of the phones dying (it all happens at once doesn't it!) I didn't bother to experiment (well I did try omreset -o off all and then omrc, but that did nothing).

I panicked and rebooted the box, and now it's good. Where should I look for what might have caused the problem?? I look in /var/opt/scalix/xx/s/logs/fatal and all i see for today (near the time of the non-responsiveness) is this: (note that even the first entry, I believe, was from me with the reset attempt, because it did pop up on the screen that POP3 wouldn't die):

SERIOUS ERROR Administration(Install Servic) Thu Jun 28 12:43:49 2007
[OM 10414] Warning. Reset incomplete. Process 3639 of
subsystem POP3 interface will not die in response to a SIGTERM signal.

Pid of logging process: 30292


SERIOUS ERROR POP3 interface(POP3 Process ) Thu Jun 28 12:45:13 2007
[OM.POP3 1026] Failed to bind the POP3 socket to a address.
Pid of logging process: 30661
Current errno value: 98


ERROR Service Router(Service Router) Thu Jun 28 12:47:32 2007
[OM 5181] Reply timed out or invalid - Mapper protocol problem.
Command sent: <none - expect greeting reply>
Reply received: 503 "ClamAV" cannot scan Scalix-owned file
Pid of logging process: 3575


ERROR Service Router(Service Router) Thu Jun 28 12:47:32 2007
[OM 5181] Reply timed out or invalid - Mapper protocol problem.
Command sent: QUIT Please Close This Session
Reply received: 220 Virus Scanning Client Ready
Pid of logging process: 3575


ERROR Service Router(Service Router) Thu Jun 28 12:47:32 2007
[OM 5183] A Mapper error has been detected.
Pid of logging process: 3575


SERIOUS ERROR Service Router(Service Router) Thu Jun 28 12:47:32 2007
[OM 5183] A Mapper error has been detected.
Pid of logging process: 3575

Would appreciate any help.

[dougp23]

Found this in /var/opt/scalix/xx/tomcat/logs:

java.net.SocketException: Too many open files
at java.net.PlainSocketImpl.socketAccept(Native Method)
at java.net.PlainSocketImpl.accept(Unknown Source)
at java.net.ServerSocket.implAccept(Unknown Source)
at java.net.ServerSocket.accept(Unknown Source)
at org.apache.jk.common.ChannelSocket.accept(ChannelSocket.java:293)
at org.apache.jk.common.ChannelSocket.acceptConnections(ChannelSocket.ja
va:647)
at org.apache.jk.common.ChannelSocket$SocketAcceptor.runIt(ChannelSocket
.java:857)
at org.apache.tomcat.util.threads.ThreadPool$ControlRunnable.run(ThreadP
ool.java:684)
at java.lang.Thread.run(Unknown Source)
Jun 28, 2007 12:22:16 PM org.apache.catalina.core.StandardWrapperValve invoke


SO I've seen the threads that say to increase the imit in /etc/security and one says to increase it in scalix-tomcat.conf. Which one is right, or do I do both? What's so bizarre is I might have had 3 users using SWA. Total users maybe 30 today.

Would appreciate any ideas! Scalix 11.0.2.23 (that's what SWA shows on login page) CentOS 4.4

[gren]

For the problem :
Reply received: 503 "ClamAV" cannot scan Scalix-owned file
see viewtopic.php?t=1333

Regards,
Gren

[gren]

There are some very worthwhile fixes in the 11.1 release and I would recommend upgrading to that. In particular, there were some problems with the message indexing code where we were not closing files when we should do under certain circumstances.

Regards,
Gren.

[dougp23]

OK, so here's a real dumb question!

I am petrified of updating. The server ran like a top for 3 or 4 months, probably 30,000 messages processed, umpteen logins, no issues. So I think I want to stay where I'm at.

That being the case, I wrote a little script that tells me how many open files java currently has. Could I reset scalix-tomcat or java and that would close the files? I am thinking my solution might be to restart these services once a month, and that would keep me healthy....

Thoughts??

[gren]

Hi,

If you stop and restart scalix-tomcat every now and again, that would work well. Of course, any clients like SWA or platform users would experience disruption whilst tomcat was down but if you're talking about once a month that doesn't sound too bad to me!

Code: Select all

/etc/init.d/scalix-tomcat stop

Check that it really has gone and wait for a bit, then.

Code: Select all

/etc/init.d/scalix-tomcat start

In theory, you should be able to use "restart" but I think sometimes that might be unreliable.

Regards,
Gren.