High CPU spikes every 1 minute

[mrcoffe]

Hi there,

Seting up Zimbra for testing a possible new email environment for the company.

I got a problem on my VM which is running Zimbra .
Every 1 minute I got a cpu spike of around 70% to 80% 24hrs a day and I don't know why. Looks lika a JAVA process is responsible for this when I watch TOP output. I would like to get rid of this!

changes that did not help:
- 'Sleep time between subsequent mailbox purges' from 1 minute to 1 hour

- fetchmail from 1 minute to 5 minutes

- disabling or editing cronjobs:
#*/2 * * * * /opt/zimbra/libexec/zmstatuslog
#*/10 * * * * /opt/zimbra/libexec/zmdisklog
#00,10,20,30,40,50 * * * * /opt/zimbra/libexec/zmlogprocess > /tmp/logprocess.out 2>&1
00,20,40 * * * * /opt/zimbra/libexec/zmlogprocess > /tmp/logprocess.out 2>&1
#0,10,20,30,40,50 * * * * /opt/zimbra/libexec/zmqueuelog
10,30,50 * * * * /opt/zimbra/libexec/zmqueuelog

- "$ zmloggerctl stop" did not help

- vi /opt/zimbra/libexec/zmlogprocess
#my $rawRetention = `${zimbra_home}/bin/zmprov -l gacf | egrep '^zimbraLogRawLifetime' | sed -e 's/zimbraLogRawLifetime: //'`;
my $rawRetention = "31d";
#my $summaryRetention = `${zimbra_home}/bin/zmprov -l gacf | egrep '^zimbraLogSummaryLifetime' | sed -e 's/zimbraLogSummaryLifetime: //'`;
my $summaryRetention = "730d";

Specs:
Using Zimbra 5.0.10
FedoraCore 7
VMware ESX Virtual machine 1 CPU 3Ghz
Nr of users: 1

Any suggestions??
Thx

[mrcoffe]

Does anybody got an idea about this problem? I still have this issue.
Thx!

[uxbod]

Which process is causing the spike ?

[mrcoffe]

It's the Java process:

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
32215 zimbra 20 0 470m 28m 12m R 43.8 2.8 0:01.54 java
25416 zimbra 20 0 689m 203m 38m S 2.0 20.1 2:35.20 java
32244 zimbra 20 0 4776 1288 1028 S 0.7 0.1 0:00.02 zmantispamctl

[kentm]

I have the same issue on a brand new install of 5.0.12 on CentOS 5.2 fully updated.

The spiking processes are Java, then Sudo, then zmmtaconfig and last about 12 seconds on a 2.4 GHz processor with 1GB RAM. They were nearly constant on a slower proc.

I have 8 users setup but I am the only one with any level of traffic (home e-mail and family calendar primarily). This occurs even when there are no users logged in.

Any insight would be appreciated.

Thanks,

Kent

[uxbod]

Requirements for a production install is 2GB+.

[kentm]

Uxbod,

The traffic on this server would probably place it in the test category by most standards but I have increased the RAM to 2GB now in any case.

I am still seeing spike to about 80% every minute, lasting for around 12 seconds.

I am using just over 820 MB RAM after the box has been running for several minutes.

Thanks,

Kent

[Perlpigeon]

Hi,
I don't have a solution unfortunately, only an experience to share.
ZCS 5.0.11 OSS, Fedora 7 x32
Specs: Intel Quad 2.4 Ghz, 2Gb RAM,
#Users: 10

Every x minutes (could be 1) the Java spikes to 20%, which I relate to a mail being offered and the spam filter rejects it. The Java process uses 21% of memory. All the other settings you mention are still on the default value.
I do recall that the logger in some version started to log in debug and that with a zmprov/zmlocalconfig command you could change the logging level. Perhaps it is located in that area?

[kentm]

Perlpigeon,

I tried turning logging off to no avail among other suggestions in the forums/wiki.

Perhaps a 20% CPU utilization on your quad is the equivalent to 80% on my single 2.4 GHz proc...

I have a catchall going to my account so if it were the AV/AS processing a message, it would end up in my Junk folder and I only get a total of about 25 e-mails per day on this home account. The other 5 users get about 25 per month combined. It is used mostly for the calendar as the kids are one week with me and one with their Mom.

There are no POP3 accounts that it goes out to check, no RSS feeds, nothing else that I can think of that it needs to do every minute. If it is checking for spam and viral updates, I can live with that once an hour or even less frequently. I guess I just need to know what it is doing and then I will try to determine how to adjust the frequency.

Thanks,

Kent

[greenieofdubbo]

Same problem here, running FC7 on VMWare ESX, Dual 2.8Ghz Xeon with 2GB Memory allocated to the Zimbra server.

Tried a fresh install with the latest version, same problem.