After a storage failure one of our Exchange databases will not mount. I ran eseutil /mh on the EDB and State says Dirty Shutdown. The log files are still on disk. What is the safe order of things to do?
Dirty Shutdown means the database was stopped before all logs were committed. If you still have the logs, soft recovery is the first and safest step.
1. Copy the EDB and all log files to another volume first. Never work on the only copy.
2. Use eseutil /mh <path to edb> and note the Log Required range.
3. Check that every log in that range is present. A missing log blocks recovery.
4. Use eseutil /r E00 /l <log folder> /d <database folder> where E00 is your log prefix. This replays the logs.
5. Use eseutil /mh again. The State should now say Clean Shutdown.
6. Mount the database.
Hard repair with eseutil /p is the last resort, because it deletes data that cannot be recovered. Do not start with it.
One log in the required range is missing. Soft recovery stops with an error.
Then look for the missing log in backups, the recycle area, or on the volume that held the old log path. If it is truly gone, restore the last good backup of the database and replay the logs you have after it. If no backup exists, copy everything first, then use eseutil /p as a last resort and expect missing items. After a hard repair, defragment with eseutil /d and start the integrity check. Many admins move the mail to a new database after that.
Found the missing log in a shadow copy. Soft recovery worked and the database mounted. Thank you.