We upgraded hardware in in the process also upgraded from RHEL4 to RHEL6. This is the master ldap of a two node setup, although both system run all services except master ldap also is logger.
/opt/zimbra lives on a SAN.
So, upgrade was based on the wiki - present LUNs to new box, install -s to install binaries, mount LUNs on new box, then reinstall (7.1.4 on 7.1.4).
Well, it didn't work quite well. LDAP choked. I had made a backup as we had similar problems when we did this in dev. However, even after the upgrade and a call to support to get LDAP back, we are still having issues:
If I log in to the node 2 admin interface, I cannot see the mail queue from node 1. If I log in to node 1 (master ldap), I can see the mail queues of both. But the worse problem is backups are not running on either system now.
On node 1, they start to run but after 5 minutes, they just stop (and I have files in /opt/zimbra/backup/tmp). A grep for zmbackup shows the following on mailbox log:
2012-06-24 01:05:58,458 INFO [btpool0-19://localhost:7071/service/admin/soap/BackupRequest] [name=spurrier@rider.edu;mid=2914;ip=127.0.0.1;ua=zmbackup/7.1.4_GA_2568;] backup - Incremental backup is started for account xxx@xxx.xxx
2012-06-24 01:05:58,458 INFO [btpool0-19://localhost:7071/service/admin/soap/BackupRequest] [name=xxx@xxx.xxx;mid=2914;ip=127.0.0.1;ua=zmbackup/7.1.4_GA_2568;] backup - Saving account information from LDAP
2012-06-24 01:05:58,459 FATAL [btpool0-19://localhost:7071/service/admin/soap/BackupRequest] [name=xxx@xxx.xxx;mid=2914;ip=127.0.0.1;ua=zmbackup/7.1.4_GA_2568;] system - out of memory
On node 2, backup doesn't even start. I only see the following entry on mailbox log:
2012-06-24 01:00:04,917 INFO [btpool0-1328://localhost:7071/service/admin/soap/AuthRequest] [ip=127.0.0.1;ua=zmbackup/7.1.4_GA_2568;] soap - AuthRequest
2012-06-24 01:00:05,183 INFO [btpool0-1328://localhost:7071/service/admin/soap/BackupRequest] [name=zimbra;ip=127.0.0.1;ua=zmbackup/7.1.4_GA_2568;] soap - BackupRequest
2012-06-24 01:00:05,184 INFO [btpool0-1328://localhost:7071/service/admin/soap/BackupRequest] [name=zimbra;ip=127.0.0.1;ua=zmbackup/7.1.4_GA_2568;] SoapEngine - handler exception
Any pointers?
I have a case with support but they won't look at it until regular business hours and we have a premium support contract - quite pissed about this..
NE 7.1.4 RHEL4 to RHEL6 issues
NE 7.1.4 RHEL4 to RHEL6 issues
Well, it seems we hit a license limit. But one box says we are within license, while the other does not. Clearly they are out of sync - But what pisses me off the most is as a paying customer with 12K accounts, I'm getting ZERO support. What's the point of having an 'access number' if you cannot get in touch with an engineer. I bet most get more help out of this forum or by googling than with support.
NE 7.1.4 RHEL4 to RHEL6 issues
You might want to talk to your sales contact about your problems with support. They might be able to get your concerns heard.
NE 7.1.4 RHEL4 to RHEL6 issues
After two business days without a decent response (I was asked to supply zmmailbox.log - there's no such thing!!!) or attempt to contact me directly, I was able to get in touch with a live rep. 2.5 hrs on the phone to get the issue escalated resulted in them having to call me back as they were switching to overseas support. 2 hours later and no call back, I FINALLY was able to get to a senior tech in India - at the same time, another senior rep/manager got involved in CA. 2 hours later, we were able the problems.
End result it was a combination of problems - ugh. Fun week.
Today we successfully migrated the 2nd server.
Solution:
* The out of memory conditions were due to a bug in the installer under RHEL6 that does not set proper number of user processes
* The backups not even starting problem on the other server was due to the directory out of sync making it show that it had more accounts than licensed
Now if someone can mark this SOLVED - cause I can't - the option is not there anywhere...
End result it was a combination of problems - ugh. Fun week.
Today we successfully migrated the 2nd server.
Solution:
* The out of memory conditions were due to a bug in the installer under RHEL6 that does not set proper number of user processes
* The backups not even starting problem on the other server was due to the directory out of sync making it show that it had more accounts than licensed
Now if someone can mark this SOLVED - cause I can't - the option is not there anywhere...