After update to 6.0 GA Statistic does not show anything

[opc31199]

After update to 6.0 GA Statistic does not show anything. Any idea.

Thank's

[EdMartin]

[My earlier posting has fallen off the radar due to its having been moved, presumably by a moderator, into a thread dealing with one of the problems I mentioned. I am hitching a ride on this thread, which apparently also concerns my second problem, in hopes of getting some assistance in solving it.]

Yesterday evening I upgraded from v5.0.18 GA FOSS to v.6.0.0 GA FOSS.

We're on a single Ubuntu 8.04 LTS 32-bit server that does nothing else but run Zimbra.

Following the upgrade, and having regenerated our self-signed certificates (because that seems to have been the only way to provoke mailboxd into life, even though the existing certificates were not in danger of expiring), Zimbra started running again.

I then noticed the following:

In the Admin console, under Monitoring | Server Statistics, the Message Count, Message Volume, Anti-Spam/Anti-Virus Activity and Disk panels display only "no data available". Yet,

* "zmcontrol status" reports that stats is running;
* I have verified that /var/log/zimbra-stats.log is accumulating data;
* I have verified that syslog is running;
* "zmstatctl status" outputs

Running: zmstat-mysql
Running: zmstat-fd
Running: zmstat-allprocs
Running: zmstat-io-x
Running: zmstat-io
Running: zmstat-cpu
Running: zmstat-mtaqueue
Running: zmstat-df
Running: zmstat-proc
Running: zmstat-vm

Perhaps it is relevant that, at my first upgrade attempt, I was informed that sysstat was MISSING, so I installed it and then the upgrade was able to proceed.

Any suggestions how I can get my server statistics displays back?

Thanks in advance for any help you can provide.

... Ed

[d2globalinc]

Originally Posted by EdMartin

Perhaps it is relevant that, at my first upgrade attempt, I was informed that sysstat was MISSING, so I installed it and then the upgrade was able to proceed.

Any suggestions how I can get my server statistics displays back?

We upgraded from 5.0.18 to 6.0 GA (Both Network Editions) - on RHEL 5.3 - We also are not seeing any statistic information yesterday after the upgrade, and now today I get notified by my admin alerts that the "stats" service has field. Sure enough I just now log into the admin console and "stats" has a nice big red "X" next to it. We have not had any issues running 5.x for the past year+, and this is the first time we have seen an issue.

I'm assuming something didn't take right with the upgrade, or there may be a slight bug in 6.0 w/ the stats? - We also had the exact same sysstats missing when doing the upgrade, so we installed it and ran through the upgrade again. We have rebooted the server a few times since the upgrade, but Its been up for over a day+ now without a restart.

We also notice that this server is consuming about 500mhz more CPU usage than it was with running 5.x from using our own stats gathering and the past month+ data. So we will be looking for a solution over the holiday weekend if anyone has an idea!

Thanks!

Shane Menshik
D2 GLOBAL INC

[pfnguyen]

Originally Posted by EdMartin

[My earlier posting has fallen off the radar due to its having been moved, presumably by a moderator, into a thread dealing with one of the problems I mentioned. I am hitching a ride on this thread, which apparently also concerns my second problem, in hopes of getting some assistance in solving it.]

Yesterday evening I upgraded from v5.0.18 GA FOSS to v.6.0.0 GA FOSS.

We're on a single Ubuntu 8.04 LTS 32-bit server that does nothing else but run Zimbra.

Following the upgrade, and having regenerated our self-signed certificates (because that seems to have been the only way to provoke mailboxd into life, even though the existing certificates were not in danger of expiring), Zimbra started running again.

I then noticed the following:

In the Admin console, under Monitoring | Server Statistics, the Message Count, Message Volume, Anti-Spam/Anti-Virus Activity and Disk panels display only "no data available". Yet,

* "zmcontrol status" reports that stats is running;
* I have verified that /var/log/zimbra-stats.log is accumulating data;
* I have verified that syslog is running;
* "zmstatctl status" outputs

Running: zmstat-mysql
Running: zmstat-fd
Running: zmstat-allprocs
Running: zmstat-io-x
Running: zmstat-io
Running: zmstat-cpu
Running: zmstat-mtaqueue
Running: zmstat-df
Running: zmstat-proc
Running: zmstat-vm

Perhaps it is relevant that, at my first upgrade attempt, I was informed that sysstat was MISSING, so I installed it and then the upgrade was able to proceed.

Any suggestions how I can get my server statistics displays back?

Thanks in advance for any help you can provide.

... Ed

Hi Ed,

Can you provide the information requested in Logger(GnR) - Zimbra :: Wiki

With this, we can identify what exactly is going on. When initially running 6.0, it may take up to a couple hours to show MTA charts (it should take well under 2 hours, in actuality).

[pfnguyen]

Originally Posted by d2globalinc

We upgraded from 5.0.18 to 6.0 GA (Both Network Editions) - on RHEL 5.3 - We also are not seeing any statistic information yesterday after the upgrade, and now today I get notified by my admin alerts that the "stats" service has field. Sure enough I just now log into the admin console and "stats" has a nice big red "X" next to it. We have not had any issues running 5.x for the past year+, and this is the first time we have seen an issue.

I'm assuming something didn't take right with the upgrade, or there may be a slight bug in 6.0 w/ the stats? - We also had the exact same sysstats missing when doing the upgrade, so we installed it and ran through the upgrade again. We have rebooted the server a few times since the upgrade, but Its been up for over a day+ now without a restart.

We also notice that this server is consuming about 500mhz more CPU usage than it was with running 5.x from using our own stats gathering and the past month+ data. So we will be looking for a solution over the holiday weekend if anyone has an idea!

Thanks!

Shane Menshik
D2 GLOBAL INC

If stats has a big red x, can you run 'zmstatctl status' to identify which zmstat component is not running?

[d2globalinc]

Originally Posted by pfnguyen

If stats has a big red x, can you run 'zmstatctl status' to identify which zmstat component is not running?

[zimbra@mx1 ~]$ zmstatctl status
Running: zmstat-fd
Running: zmstat-proc
Running: zmstat-mysql
process 10116 in /opt/zimbra/zmstat/pid/zmstat-allprocs.pid not running
Running: zmstat-mtaqueue
Running: zmstat-vm
Running: zmstat-io-x
process 10120 in /opt/zimbra/zmstat/pid/zmstat-convertd.pid not running
Running: zmstat-io
Running: zmstat-cpu
Running: zmstat-df
[zimbra@mx1 ~]$

Was the response from 'zmstatctl status' - and I have also included a screenshot I've taken of the CPU usage of the Zimbra VM. You can see the changes after the upgrade around 2AM Thursday. Not sure if its related or not, but just thought I would post some info to backup what I was stating in the previous message.

Let me know if you need any-more information, if this can help out the community I'm all for it! We are enjoying Zimbra 6 - and I'm putting together a list of our other findings / issues and will compare those to open bug reports as well and post in a new thread.

I will be around working this holiday weekend and will be more than happy to perform any test or gather any information if needed.

Thanks again!

Shane Menshik
D2 GLOBAL INC

[pfnguyen]

Originally Posted by d2globalinc

[zimbra@mx1 ~]$ zmstatctl status
Running: zmstat-fd
Running: zmstat-proc
Running: zmstat-mysql
process 10116 in /opt/zimbra/zmstat/pid/zmstat-allprocs.pid not running
Running: zmstat-mtaqueue
Running: zmstat-vm
Running: zmstat-io-x
process 10120 in /opt/zimbra/zmstat/pid/zmstat-convertd.pid not running
Running: zmstat-io
Running: zmstat-cpu
Running: zmstat-df
[zimbra@mx1 ~]$

Was the response from 'zmstatctl status' - and I have also included a screenshot I've taken of the CPU usage of the Zimbra VM. You can see the changes after the upgrade around 2AM Thursday. Not sure if its related or not, but just thought I would post some info to backup what I was stating in the previous message.

Let me know if you need any-more information, if this can help out the community I'm all for it! We are enjoying Zimbra 6 - and I'm putting together a list of our other findings / issues and will compare those to open bug reports as well and post in a new thread.

I will be around working this holiday weekend and will be more than happy to perform any test or gather any information if needed.

Thanks again!

Shane Menshik
D2 GLOBAL INC

Can you copy the output of /opt/zimbra/zmstat/zmstat.out to determine why the two zmstat collectors are not running? (I will be on vacation all of next week and back the following Tuesday)

[d2globalinc]

Originally Posted by pfnguyen

Can you copy the output of /opt/zimbra/zmstat/zmstat.out to determine why the two zmstat collectors are not running? (I will be on vacation all of next week and back the following Tuesday)

[zimbra@mx1 zmstat]$ cat zmstat.out
09/04/2009 04:52:00: ERROR: No such process: '1019'
09/04/2009 05:31:30: ERROR: No such process: '1259'
09/04/2009 14:45:30: ERROR: No such process: '13686'
09/04/2009 14:45:30: ERROR: No such process: '13687'
[zimbra@mx1 zmstat]$

and

[zimbra@mx1 zmstat]$ zmcontrol status
Host mx1.d2gnet.com
        antispam                Running
        antivirus               Running
        convertd                Running
        ldap                    Running
        logger                  Running
        mailbox                 Running
        mta                     Running
        snmp                    Running
        spell                   Running
        stats                   Stopped
[zimbra@mx1 zmstat]$

[d2globalinc]

Original email alert I received from Zimbra -

Sep  3 23:45:05 mx1 zimbramon[4575]: 4575:err: Service status change: mx1.d2gnet.com stats changed from running to stopped

[pfnguyen]

Originally Posted by d2globalinc

[zimbra@mx1 zmstat]$ cat zmstat.out
09/04/2009 04:52:00: ERROR: No such process: '1019'
09/04/2009 05:31:30: ERROR: No such process: '1259'
09/04/2009 14:45:30: ERROR: No such process: '13686'
09/04/2009 14:45:30: ERROR: No such process: '13687'
[zimbra@mx1 zmstat]$

and

[zimbra@mx1 zmstat]$ zmcontrol status
Host mx1.d2gnet.com
        antispam                Running
        antivirus               Running
        convertd                Running
        ldap                    Running
        logger                  Running
        mailbox                 Running
        mta                     Running
        snmp                    Running
        spell                   Running
        stats                   Stopped
[zimbra@mx1 zmstat]$

I don't know why they would not run for you; can you run the following from the commandline:

1. /opt/zimbra/libexec/zmstat-allprocs -c -i 5 # CTRL-C after a dozen seconds if it runs
2. /opt/zimbra/libexec/zmstat-convertd -c -i 5 # do the same as above