zmlogprocess - "Out of memory!"

[xtrac]

Hi Folks,

I am reclaiming some disk space from the logger, its currently at 18GB with the default Lifetime values. Running on ZCS 5.0.11 OSE on Debian 4.0, with store of about 88GB.

I have changed them now to be the following

==========
zimbra@faraday:~$ zmprov mcf zimbraLogRawLifetime 7d
zimbra@faraday:~$ zmprov mcf zimbraLogSummaryLifetime 365d
==========

After this I ran

==========
zimbra@faraday:~$ ./libexec/zmlogprocess
==========

Which gave me the following output.

==================
Pruning service/server entries older then 2009-02-23 23:05:02
faraday maps to faraday.XXXX.XXX
Processing mta entries from 134136159
Found 691958 messages
Inserted 691958 rows
Saving ID 222411382
Deleting processed postfix logs from raw_logs...Done
Processing clamd entries from 94474308
Found 0 entries
Deleting processed clamd logs from raw_logs...Done
Processing amavis entries from 134140318
Out of memory!
==================

Its still running, its been running now for about 19hours. The logger directory is still 18GB.

How long should I leave it run after this notice? Is there a parameter I can set for zmlogprocess to be able to use more memory? Or, should I try for larger day values first, then work down to smaller ones? (Dont want to do this if necessary, as it is clearly taking a significant amount of time.) Or, can I just ignore the error and wait for it to finish?

Regards

-Roy

[xtrac]

Well,

I've let it finish, it finished shortly after it gave the out of memory error.

The following processes are still running.
=============
faraday:/opt/zimbra# ps -ef | grep zmlogprocess
zimbra 30243 30238 0 16:51 ? 00:00:00 /bin/sh -c /opt/zimbra/libexec/zmlogprocess > /tmp/logprocess.out 2>&1
zimbra 30288 30243 0 16:51 ? 00:00:57 /usr/bin/perl /opt/zimbra/libexec/zmlogprocess
root 7249 22844 0 18:48 pts/1 00:00:00 grep zmlogprocess
faraday:/opt/zimbra#
=============

However the logger directory is still 18GB. Any pointers or ideas here would be most welcome.

Regards,

-Roy

[phoenix]

There was a problem with the logger process causing OOM errors, you should upgrade to the most recent version of Zimbra. As a temporary measure you can update the script in comment #9 of the bug report.

[xtrac]

Thanks Bill,

I'll upgrade at the weekend. zmlogprocess hosed the machine trying to clear out the logger db, so I just deleted them and started it from scratch, using the instructions from King0770-Notes.

Regards

-Roy