Solved

What happens when a mailbox archive job doesn't finish a mailbox?

  • 3 August 2022
  • 4 replies
  • 33 views

Badge +1

This is probably a really basic question but I just need validation that this is working the way I suspect it is.

I have some huge Exchange mailboxes, some of them containing over 3 million messages.  I also set limits on how long my mailbox archive jobs can run so they are not active during the day and adding load onto our Exchange system.

My question is, if a mailbox archive job is stopped before finishing does the next job treat the messages that were done as completed?  For example, a mailbox has 100,000 items in it and the archive Monday night completes 50,000 of them before being stopped.  Does the next job try to archive all 100,000 messages again or just the 50,000 that were not finished?

icon

Best answer by Scott Reynolds 4 August 2022, 16:56

View original

4 replies

Badge +4

The mailbox archive job reads every mail item in the inbox, each time it runs. It applies the archive rules to any item that meets the requirement. The archive process moves item to the Commvault archive, writes a stub to mail box, and deletes original mail item. Once an item has been archived it moves to the next one in the mailbox.

 

So in answer to your question a failed job does not mean all those archived items are lost. The bad news is that if you are replacing an archived mail with a stub the number of items in the mailbox will not change, so the next archive operation still has the read the same number of mails, and this is what takes time. Big mailboxes take ages to process.

 

 

Badge +1

Thanks for the reply, but I’m actually referring to the backup process not the stubbing process.  So to clarify my original question: If a mailbox contains 100,000 items and the backup stops after processing 50,000 of them, does the next backup job reprocess the entire 100,000 or only the 50,000 that were not completed during the first job.

Userlevel 5
Badge +11

Hello @David Lewis 

No the backup will commit what has already been processed. For any reason the backup is killed/failed etc anything that has already been processed is retained in the backup and NOT processed again in the next backup.

This is helpful when you have very large mailboxes, etc.

An example here for those very large mailboxes, you can set a total run time on the backup job so that it's killed at X time. Whatever was processed during that job is retained and next backup will start where that previous job left off.

 

Badge +1

Thank you, that is what I suspected, but these mailboxes are so big they had me second guessing myself.

Reply