Earlier this morning, I saw the following message on the console of one of our Domino servers.
Error reading OperationTime for folder operations in database d:\Programme\IBM\Lotus\Domino\data\mail\xxxx.nsf. Run Fixup on the database.: Invalid Time or Date Encountered
The database belongs to a user that have quit the company a few days ago and I was not very concerned about this message. As usual I would make a new copy of the database and store it on a different server and let adminp do the job of deleting the user and the according database and replicas.
I created a new copy and opened the database. A messagebox appeared:
Folder has been damaged. Please close and reopen database to have it repaired.
Needless to say that a fixup did not solve the problem. Nice to know that “a folder” is damaged, but why does the message does not tell me which folder is damaged?
My only hope was, that the server log would be more verbose. So I opened the log. And there it was:
Database marked corrupt: Folder ($Inbox) corrupt
06.08.2009 11:07:33 Database collection error: d:\Programme\IBM\Lotus\Domino\data\xout\xxxxx.nsf: Folder has been damaged. Please close and reopen database to have it repaired.
The strange thing is that the database on the source server opens without any error message…
I deleted the copy from the server and ran a fixup against the source database. Then I replaced the design and deleted all view indexes.
Once again I created a new copy … same error occured.
I opened the source database and moved all documents from the inbox to a temporary folder. I deleted the inbox and replaced the design to recreate the inbox folder
Guess what happened after the new copy had been created .
I created a new database on the source server just to be sure that it is not the template that is damaged. The new copy on the target server opens without any complaint.
Next thing I tried is to create a new copy on the target server after deleting the inbox and before replacing the design. I replaced the design after the database was copied to the target.
And … drumroll … No error message in the client or in the server’s log.
Can anyone explain this behaviour, please??
I’ve seen this problem with a mailfile after a problem on an 8.5 server where nhttp.exe had locked the database for access somehow (browser/client/server task access attempts resulted in the ‘database is in use by another … ‘ message. Stopping the http task released the lock but that’s when I started seeing this error for this particular database.
Can’t offer further insights though as to why and how to fix/prevent it.
Thx for the hint. Will do some further investigation.
We’ve seen this several times when users set the date on their computers into the future.
@Andrew That was my first thought. But when I checked the workstation, everything was OK. By the way, how far is “into the future” ? Could a few minutes time difference also cause the problem ?
Reminds me of this regression: http://www-01.ibm.com/support/docview.wss?uid=swg21290840
in case there is a single server and databse gets corrupted then get it restored from a healthy backup.