zmamavisdctl stopped

[lsu_guy]

Hi Everyone,

Last night, a service (zmamadvistctl) stopped on our production server. It then stops delivering any messages. I restarted this service using:
> zmamavisdctl start

This started the service but bogs down the server to the extent that I cant even SSH into it. I started it once this morning and it died 10 minutes later.

I was finally able to get through to it using ssh, and restarting zimbra through "zmcontrol stop" and "zmcontrol start" seems to have fixed the issue. (for now).

Is there a reason that this service would stop all emails from being received or delivered to individual mailboxes?

[brian]

I would take a look at /var/log/zimbra.log for any amavis error messages that would indicate why it stopped or crashed.

[dwmtractor]

Might this be another instance of what I posted here?

Antivirus failing to start in scripted restart

[lsu_guy]

I'm going through the log and some of the things that jump out at me are:

Aug 3 22:34:01 <host> amavis[29010]: (29010-04) (!!)TROUBLE in check_mail: parts_decode_ext FAILED: file(1) utility (/usr/bin/file) error: run_command (open pipe): Can't fork at /opt/zimbra/zimbramon/lib/i486-linux-gnu-thread-multi/IO/File.pm line 188. at /opt/zimbra/amavisd/sbin/amavisd line 2603.
Aug 3 22:34:01 <host> amavis[29010]: (29010-04) (!)PRESERVING EVIDENCE in /opt/zimbra/data/amavisd/tmp/amavis-20090803T223042-29010
Aug 3 22:39:53 <host> amavis[5814]: (05814-20) (!)lookup_ldap: timed out at (eval 73) line 138, <GEN196> line 5.
Aug 3 22:41:21 <host> amavis[5814]: (05814-20) (!)Requesting process rundown, task exceeded allowed time
Aug 3 22:41:22 <host> zimbramon[11220]: 11220:info: zmstatuslog timeout after 60 seconds
Aug 3 22:41:23 <host> zimbramon[11203]: 11203:info: zmstatuslog timeout after 60 seconds
Aug 3 22:57:44 <host> slapd[28459]: => bdb_idl_delete_key: c_del id failed: DB_LOCK_DEADLOCK: Locker killed to resolve a deadlock (-30995)

[lsu_guy]

I think the amavis error is what caused this problem.

I found someone else who had a similar issue (this is a non-zimbra server).

Can anyone tell me the zimbra way to fix this problem?

[lsu_guy]

Dan...I dont think this problem is the same as the one you posted.

Can anyone give me any direction on this? I would like to get this solved.

For now, I created a zimbra monitoring service on another server which queries the status of my servers and emails/txts me whenever a service stops running. I run this service every 15 minutes and I also use an external delivery smtp (google) to send me those error messages. Although this is nice to have, it still doesnt explain what happened to our servers a few days ago. If anyone else is interested in monitoring their server this way, I can add them to my monitoring service.

[lsu_guy]

Is there anyone from zimbra who can reply to this? Any feedback is appreciated. We have ZCS NE.

[phoenix]

Search the forums for 'parts_decode_ext FAILED fork', see if the solved solution fixes your problem.

[lsu_guy]

Hi Bill...I looked at the following links:

1) [SOLVED] parts_decode_ext error
Errors in admin mail report

I have all three of those packages installed.


2) Fork error

This thread seems unresolved



I think this line "amavis[5814]: (05814-20) (!)Requesting process rundown, task exceeded allowed time" is where amavis actually died.

So far, after the restart it seems to be running fine.

[sahil]

Hi,

I am using Zimbra, version of my zimbra is Release 5.0.9_GA_2533.RHEL5_20080815132719 RHEL5 FOSS edition and when I am checking zimbra zimbra service status showing
[zimbra@test ~]$ zmcontrol status
Host test.mydomain.com
antispam Stopped
zmamavisdctl is not running
antivirus Stopped
zmamavisdctl is not running
ldap Running
logger Running
mailbox Running
mta Running
snmp Running
spell Running
stats Running

I have start stop Zimbra service's and even restsrta Zimbra but no benefit.