Yes, /opt/zimbra/store (not data), sorry about that.
I pulled a backup image of the machine from the day before and it was in the same state (only files from Dec 28 and earlier in store/). I'm going to grab another image from the last known running state.
But I believe I have found out what happened. We had some issues a few months ago when the there was a power outage and the dbm files were corrupt (LDAP). I removed the lock/tmp files and things sprang back to life. It looks like a system admin did the same thing this weekend to try and recover from this weekends power outage.
Looks like I'll need to pull the new emails over from one image into the other to keep things running. Can I just copy over the store files and re-index? |