MySQL.Server Fails (Cannot Send Mail)

[msf004]

Currently, and seemingly suddenly, the MySQL.Server instance will not remain running on my server - or is at least having difficulties.

Also, oddly enough, I can receive mail, but cannot send mail.

If I do a ZMControl stop/start, then look at the status, all is fine; however, as soon as someone attempts to SEND an email, the server hangs - after which, if I attempt a zmcontrol status the status polling timesout after logger...running (the next test is mailbox).

If I can get a zmcontrol status to respond, it then returns MySQL.server is not running.

I can try a mysql.server start, but I get a result of "mysql server is already running ... blah blah blah.

Anyone have any kind suggestions?

Server has been running fine for months upon months with no issues.

[veronica]

Any errors in /var/log/zimbra.log when you send ?

[stiller]

Is mysqld actually running? (ps aux | grep mysqld)
Can you connect to mysql? (run mysql as user zimbra)

[msf004]

I see a few things:

zmtomcatmgr[14979]: file /opt/zimbra/log/zmtomcatmgr.pid does not exist
zmtomcatmgr[14979]: assuming no other instance is running
zmtomcatmgr[14979]: no manager process is running

amavis[17384]: Net::Server: Group Not Defined. Defaulting to EGID '501 502 501 5 4'
amavis[17384]: Net::Server: User Not Defined. Defaulting to EUID '502'

I am also seeing this in the mailbox.log:

WARN [Timer-0] [] dbconn - Connection pool is 75% utilized. 19 connections out of a maximum of 25 in use. Turn on debug logging for zimbra.dbconn to see stack traces of connections not returned to the pool.
2009-12-20 11:29:13,074 WARN [Timer-0] [] dbconn - Connection pool is 75% utilized. 19 connections out of a maximum of 25 in use.

...and I know I corrected that awhile back, and of course I forgot how I corrected it and where I increased the connection pool.

[msf004]

Stiller, per the end of the previous reply, when I run mysql I get:

too many connections

[stiller]

You can try increasing the max_connections setting in /etc/my.cnf

[msf004]

I increased the max_connections in my.cnf and I increase the zimbra_mysql_connector_maxActive setting.

However, there are still some odd issues.

I have noticed that when I poll for status, whether via the web admin console or via zmcontrol status via a terminal, the whole machine just stalls.

I wonder if this this a hardware issue? Like I said above, everything was fine up until 8 AM yesterday.

[msf004]

This is the consistant result for a status:

./~] # zmcontrol status
Host server.domain.net
antispam Running
antivirus Running
ldap Running
logger Running
Timeout after 180 seconds