New ZCS 6.0 on Debian 5 Having Webmail '500' Errors

[ndconsulting]

Hi everyone,
I hope you guys can help me out on this one. Im new to Zimbra and I just completed a install of ZCS 6 on a brand new Debian 5 box. Its running on a 4gb slice server at SliceHost.com.

The install went smooth and mail is running, but about every 8 hours, the Webmail becomes unavailable. When you load it in a browser, about 20% of the time the user will get what i'm assuming is a jetty generated error message saying:

'Error 500 - An Internal Server Error Occured'.

The other 80% of the time, the browser completes the load and its a completely blank screen. Nothing there at all, no errors, no html.

The server appears ok after this happens. CPU and load averages are low, memory usage seems normal.

The only way I have been able to recover from the error is to do a full stop and start of the zimbra services. After that, it works great for another 8 hours or so.

My first thought was that the server didnt have enough memory, so I upgraded it to a full 4gb, but no luck.

Im brand new to zimbra to I read through the forums and the admin guide and nothing seems to standout on my config that would cause it. I have checked most of the zimbra log files and nothing seems to stand out there either, but im not exactly sure where to start looking.

Here the edition Im running:
Release 6.0.0_GA_1802.DEBIAN5_64 DEBIAN5_64 FOSS edition.

Can someone point me where to start checking or what logs to examine closer?

Thanks!

[ndconsulting]

Anyone have any ideas? I just had to restart all the zimbra services again. It made it a full 24 hours this time, but this morning both webmail and imap proxy were down. Nobody was able to check mail via IMAP or webmail until all services were restarted.

I did notice that after I stopped all the zimbra services, the machine was still using ~2.7G of memory and it only released ~1.3G.

Any ideas or direction?

[ndconsulting]

ANother update.

When restarting the services this time. This is the result I get.

Starting logger...Done.
Starting mailbox...Done.
Starting imapproxy...Done.
Starting antispam...Done.
Starting antivirus...Done.
Starting snmp...Done.
Starting spell...Failed.
Starting apache...(98)Address already in use: make_sock: could not bind to address [::]:7780
(98)Address already in use: make_sock: could not bind to address 0.0.0.0:7780
no listening sockets available, shutting down
Unable to open logs
failed.


Starting mta...Failed.
Starting zmmtaconfig...zmmtaconfig is already running.
postfix failed to start
Starting saslauthd...done.


Starting stats...Done.



This was after doing a full reboot of the server and then starting the zimbra services using 'zmcontrol start'

Any ideas?

[ndconsulting]

Just wanted to update my thread. Im still having massive issues with Zimbra. Its almost unusable at this point. About every 8 hours depending on usage, all the zimbra services have to be restarted.

When it stops, both IMAP and Webmail stop functioning. On the last crash, I killed the java process and webmail came backup, but no IMAP access.

I have checked the logs and I cant find anything out of the ordinary. I have set the max IMAP connections to 1500 thinking that maybe it was a connection limit issue.

When it stops responding, the sever itself is running fine. Memory usage appears to be good. Just no IMAP or webmail connections at all.

Any ideas or can someone point me in a direction to start looking. I have been reading everything I can in the wiki and forums with no luck.

Any ideas?

[ndconsulting]

I have attached the stack trace from the last time it stopped responding. Maybe that will help.