cancel
Showing results for 
Search instead for 
Did you mean: 

Give Exchange Mailbox Archiving Task More Threads?

krismason
Level 3
Hi All, It seems that our Archiving task is not utilising all of the threads that it could. Is it possible to increase the amount of threads it is using? We are using Exchange 2003 with EV 8 SP4 all virtualised. Thanks Kris
1 ACCEPTED SOLUTION

Accepted Solutions

Rob_Wilcox1
Level 6
Partner
The optimum is different in different environments as you would expect.  There are a lot of variables, as you would also expect.  If you read the EV 8 Performance Guide that will give you an indication of some of the environmental details that can be used in your "equation".

Hope that helps,
Working for cloudficient.com

View solution in original post

4 REPLIES 4

Kelley_Allen
Level 3

You can edit the properties of the Exchange archive mailbox task and set the concurrent connections to the exchange server higher.

However, what I found out is that the optimum setting for our enviornment is 6 with maximum number of items per pass set to 200.

This seems low, but I did verify with Symantec support that less is more in this case.  Having these settings in this range will bascially archive 6 mailboxes at the same time and archive 200 messages with each run.  I did however, when I first implemented the archive server set the maximum number of items per pass to 500 and ran the task on specific larger mailboxes individually.


krismason
Level 3

Hey Kelley,

Thanks for the response. Its currently set at 5 and 1000.

I think now we have caught up on our backlog we will leave it at this.

Thanks again

Kris

Rob_Wilcox1
Level 6
Partner
The optimum is different in different environments as you would expect.  There are a lot of variables, as you would also expect.  If you read the EV 8 Performance Guide that will give you an indication of some of the environmental details that can be used in your "equation".

Hope that helps,
Working for cloudficient.com

JesusWept3
Level 6
Partner Accredited Certified
yeah and with some of the mapi leaks out there, more threads and less per pass would mean the leak would be more prominent and the more likely you would run in to Mutex issues and the like, but i guess my question would be how are you determining it wasnt using all the threads? purely because there was a backlog?
https://www.linkedin.com/in/alex-allen-turl-07370146