cancel
Showing results for 
Search instead for 
Did you mean: 

Possible to run Move Archive from another EV server?

EVGuy
Level 4

Hi All,

I'm trying to figure out if it's possible to run Move Archive from another server. What I have in place is one EV server doing mailbox archival. I'd like to bring up another EV server to join the site and run Move Archive from that server. The reason behind this is that the moves are process intensive and I'd like to have a more powerful server (which also is doing nothing else) doing the moves for me and leave the original server to just continue it's archiving.

Is this possible? 

Also can anyone confirm that a users archived mail is 'read only' while the move is being done to their archive? I found mention of it in the documentation but am not sure what would happen if a user actually tried to retrieve the message from the stub while the move was in process.

 

Thanks!

Mike

1 ACCEPTED SOLUTION

Accepted Solutions

JesusWept3
Level 6
Partner Accredited Certified
Move archive tasks are really tied to the storage service So if you have two EV servers both with vault stores, and you were logic all your users to another site, you would find that you would have two move archive tasks created and each one would process users that are hosted on that vault store located on that server If you had a much more powerful machine, you could move the vault stores and storage service to that newer machine and the task will run on there
https://www.linkedin.com/in/alex-allen-turl-07370146

View solution in original post

8 REPLIES 8

EVGuy
Level 4

Bueller? Bueller? Anyone?  :)

Can this not be done? I can't imagine having an EV site with multiple servers and you can't control who does all the processing when running 'move archive'

JesusWept3
Level 6
Partner Accredited Certified
Move archive tasks are really tied to the storage service So if you have two EV servers both with vault stores, and you were logic all your users to another site, you would find that you would have two move archive tasks created and each one would process users that are hosted on that vault store located on that server If you had a much more powerful machine, you could move the vault stores and storage service to that newer machine and the task will run on there
https://www.linkedin.com/in/alex-allen-turl-07370146

Rob_Wilcox1
Level 6
Partner

JW2 is right, it's really going to be as intensive/painful as how fast/slow your storage server is.

Working for cloudficient.com

EVGuy
Level 4

Is this going to have a huge impact on users or on the mailbox archive process if Move Archive is running while these things are going on? I've moved a few test users and the CPU and memory usage while running Move Archive are quiet significant. 

I guess my goal here is to reduce the impact to the EV users while running the Move. (Which was why I wanted to have a third machine that was doing nothing else doing the heavy lifting.) 

:)

Rob_Wilcox1
Level 6
Partner

That's a "how long is a piece of string" type of question, I'm afraid.

 

However your archiving task will be running out of hours no?  What window do you have defined?  What window do you have for your Move Archive task?

Working for cloudficient.com

EVGuy
Level 4

I'd have to check to see exactly how long the mailbox archive tasks takes to complete. I'm just wondering if there's any issue if there's some overlap.

The hardware that the EV server is on spikes at about 90% CPU when running a move on one test account. However I've only tried it when I knew the Mailbox Archive process wasn't running.

The other thing is if is overlap or not, does it affect my end user when the Move is running? I found in the document that the archive is in a 'read only' state. But does that mean they can only read the stub information in the mailbox or does it mean that the message can be recalled but new messages cannot be archived into the vault while the move is running?

 

Rob_Wilcox1
Level 6
Partner

IMHO some overlap is fine, but best, as with many things, to try to keep them seperate.  90% CPU usage isn't uncommon, but remember Windows will multitask away, so it's not like you've "only" got 10% CPU left for everything else that the system must do.

 

The archive will be readonly, meaning that they can recall/retrieve items, but they can not archive new data.

Working for cloudficient.com

AndrewB
Moderator
Moderator
Partner    VIP    Accredited

how many archives and how much data are you talking about here? i have another idea that would fit your use case of having a 3rd server do all the work.