Watchfile Errors after a Partition move

[Scott_Lewis_2]

I recently moved a partition using the method described in http://support.veritas.com/docs/273271.  I then closed the partition and opened a new one.  Everything worked okay but now when storage file watch runs I am filling up my event logs with errors (7109).  In reviewing the SQL tables, I have found 4100 entries in the Watchfile table that has the previous location hardcoded in the ItemName column.  I have tried running the Archive Task in Report mode with the Pending Shortcut Timeout set to “0” but that did not work.  Any ideas on how I can clear the Watchfile table and/or clear the errors?
 

[RahulG]

Starting with EV 8 SP1, when Enterprise Vault comes out of backup mode, the StorageFileWatch process identifies that the path is no longer valid and the archiving operation of these files will attempt to be canceled if your Vault Stores are set to remove safety copies "After Backup".  If it's successful, the items will be removed from the following locations;
- Vault Store partitions.
- Indexes.
- Vault Store and Fingerprint databases.
Additionally, the items residing in the user's mailboxes will be canceled from a Pending Archive item, back to it's native format thus giving it another chance at being archived.

If your Vault Stores are set to remove safety copies "Immediately after Archiving", the items will be successfully removed from the WatchFile table when Enterprise Vault comes out of backup mode.  Additionally, the items are not duplicated in the index or in the vault stores and will remain an archived item in the user's mailbox.

Because of this self healing feature, this event warning can be safely ignored

[Scott_Lewis_2]

Thanks for the feedback – unfortunately what you are describing is not working.  Since I have moved the archive to a different partition the Watchfile table never gets updated or removes the items.  I have the store set to remove safety copies "Immediately after Archiving" and but the store in and out of backup mode but all I get is 4000 errors that the file cannot be found:

Partition Name = APArchive Ptn3
 VaultStoreEntryId = 1E3D2076AA8430A4DAFB8A5B467E7979A1210000apsstcev.ap.net
 TransactionID = B0BAC1A9664220B3ABDA6C91B5E0F291
 Invalid Reason = File 'E:\EVault\data3\2010\07-24\B\0BA\B0BAC1A9664220B3ABDA6C91B5E0F291.DVS' not found for SavesetIdentity = [7209930]
 
I have tried changing the store to “after backup” and stopping and starting the storage service but it still does not make a difference.  Does anyone know what are the ramifications if I were to delete the entries from the Watchfile table?
 

[RahulG]

Does the dvs file which is mentioned in the event exist on the location?

[Scott_Lewis_2]

The partition was move to the D: drive which had more space. So yes, it does exist on the D: drive but no longer on the E: drive.  Sounds like from your eariler post the watchfile entries should have been updated to the new location after I changed the location in the partitionentry table

[Batmanfail]

Your DBA should be able to directly update that table with the correct path.  This KB article explains it;  http://seer.entsupport.symantec.com/docs/287955.htm 

Something like;

USE VaultStoreDB
GO
UPDATE dbo.WatchFile
SET ItemName = N'D:\New Partition\'
WHERE ItemName = N'E:\EVault\data3\'
GO