[SOLVED] After install server status shows errors for all services

[Oded Arbel]

I've just finished installing ZCS 6.0.4 on a CentOS5 EC2 instance. I had to try multiple times until I got the exact IP setup working, but now it looks to be working correctly - e-mails are going in and out, and I also tested calendar, documents and tasks and I don't seem to have any problems, but the server status screen shows red X marks for all the services on my single server.

Could there still be problems with the IP setup? The current IP setup I use is that the server resolves its own hostname (which is something like ec2.mydomain.com) to the EC2 internal IP address and I installed it with that domain (it complained a bit about not having an MX record but I ignored it). I've set the admin address to 'admin@mydomain.com' and it automatically created the domain 'mydomain.com' in the domains list, in addition to 'ec2.mydomain.com'.

I also looked at all the logs in /opt/zimbra/log, and there is a lot of info there so I might have missed something, but there were no obvious errors that I found.

As far as I can tell everything is working correctly except the red X marks in the server status screen. Is there some way I can tell what is the real problem?

[uxbod]

Welcome to the forums

Logger (ZCS 6.0.x and later) - Zimbra :: Wiki

[Oded Arbel]

Thanks. I looked under "Server Statistics" and got this error message:
Server error encountered Message: system failure: Unable to read logger stats Error code: service.FAILURE Method: GetLoggerStatsRequest Details:soap:Receiver

I then looked in the trouble shooting section of the page you mentioned and tried 'zmloggerctl status' and got no response. with 'start' it claimed that it is already running and a 'restart' produced to OK messages and went back to the same (non)status.

Running 'zmsoap -z GetLoggerStatsRequest stats/@name=zmmtastats' returned:
ERROR: service.FAILURE (system failure: Unable to read logger stats)

While running '/opt/zimbra/libexec/zmrrdfetch -f zmmtastats' got me:
DBD::SQLite::db prepare failed: no such table: rrds at /opt/zimbra/libexec/zmrrdfetch line 43.
Can't call method "bind_param" on an undefined value at /opt/zimbra/libexec/zmrrdfetch line 48.

What do I do now?

[uxbod]

Admin console server status page show stopped services

[simonfishley]

It's all very well linking a user to another post but generally one hopes the post will actually contain an answer. In this case it does not. In the other post you suggest

"Search the forums for SSH (zmsshkeygen) ... it sounds like that ZCS can ssh to the zimbra user to get the services status."

Which is quite cryptic. The other user tried to regen the SSH keys, as have I and it has not helped.

It seems this is avery common problem and while it does not affect prpoer functioning of the server, it is worrying not to be able to have a quick look to see all the services are up.

Any other suggestions we may try?

Thanks
Simon

[Oded Arbel]

I think the problem is unlikely to be SSH related, as I don't have any remote servers - only 1 local server. The logger service is obviously at fault as it doesn't appear to be working on my site and generates SQLite errors when trying to access the RRD.

I'm not familiar with the logger service, and the wiki page uxbod pointed to plainly says to report this in the forums, so here I am reporting ;-)

[phoenix]

Originally Posted by Oded Arbel

I'm not familiar with the logger service, and the wiki page uxbod pointed to plainly says to report this in the forums, so here I am reporting ;-)

The Troubleshooting section on the wiki page actually says this:

Take the above information, and post/attach the information to support or the Zimbra Forums

You've not actually provided any information asked for on that page.

[Oded Arbel]

I kind of did - see my second message above

[phoenix]

Originally Posted by Oded Arbel

I kind of did - see my second message above

Yes, 'kind of' - that was part of the information and the rest of it is also useful for debugging.

You should also post the output of the following commands (run on the zimbra server):

cat /etc/hosts
cat /etc/resolv.conf
dig yourdomain.com mx
dig yourdomain.com any
host `hostname` <-- use that exact command with backticks not single quotes

[Oded Arbel]

Thanks for your response.

On the face of it, the information you requested doesn't seem to be related to the logger (that I now believe to be the problem) and none of these things are mentioned in the Wiki page referenced.

That being said, here is the data (removing some redundant stuff like diggin MX records and then digging any):

# cat /etc/hosts
127.0.0.1 localhost localhost.localdomain
10.228.102.117 vhost1.heptagon.co.il vhost1 heptagon.co.il Coming Soon

# cat /etc/resolv.conf
search eu-west-1.compute.internal
nameserver 172.16.0.23

# dig heptagon.co.il any
;; ANSWER SECTION:
heptagon.co.il. 46098 IN NS ns.eonspace.net.
heptagon.co.il. 46098 IN NS ns2.eonspace.net.
heptagon.co.il. 35765 IN MX 10 mail.heptagon.co.il.
heptagon.co.il. 35765 IN MX 20 mail.eonspace.net.
heptagon.co.il. 35764 IN A 79.125.24.251

# host `hostname`
vhost1.heptagon.co.il has address 10.228.102.117

Thanks in advance for any input that you have on the problem. I think the main issue is the problem that can be seen by running:
# /opt/zimbra/libexec/zmrrdfetch -f zmmtastats
DBD::SQLite::db prepare failed: no such table: rrds at /opt/zimbra/libexec/zmrrdfetch line 43.
Can't call method "bind_param" on an undefined value at /opt/zimbra/libexec/zmrrdfetch line 48.

I'm pretty sure it should be fairly easy to fix this problem - which has a very good reporting including line numbers - to anyone familiar with that tool.

Best regards