This server was run on dmraid centos 5.4. On this server we had kernel crush because of problems to reinitialize the dmriad. Mail store was contain 15 internal mailboxes with emails.
After crush was need reboot the server. I logged into recovery mode and copy out the zimbra folder in /opt directory. Unfortunately the raid error was unrecoverable. System was re installed to Centos 6 x86_64.
To recover zimbra I did copy of the forder and copied to /opt directory and downloaded last version of zimbra, unpacked and started installation and it asked me if I want to upgrade and I answered "yes". All went through after second attempt because I for got clean SSL certificate and it was complain that certificate not match host name. After that all ran no error except that centos 6 use rsyslog and some reason zimbra stats logged data into both logs message and zimbra-stats. Zimbra running, web client working, and no problem with email flow. But old emails left in store directory I can see them when doing the cat on *.msg.
I red documentations about Disaster Recovery. Also few people suggested use zexstras to perform organized backup. I installed for right now trial version and it running each morning to perform backups on NFS mount.
I tried run some re index command like:
Quote:
510 zmprov rim email address start
511 zmprov rim email address status |
But it not really help and I want to know right way to recover old emails.