Tasks...scheduling...and real world stuff

[Micah_Wyenn_2]

here's a q about the scheduling of tasks. Let's say I've got my mailbox archiving task set to run from midnight to 6am. I have one enabled mailbox. At midnight it runs and archives my mailbox.

So what happens next to my task...since it's completed all the mailboxes in my enabled list, does it just go to sleep until the next midnight...or does it leap back in to my mailbox and analyze it to see if there's more stuff to archive.

What happens if I add another enabled mailbox during the operational window, does the task see that and dive back in to the mailbox?

What happens if the enabled mailbox is huuuuuge, and the end time of the task comes during it's processing...does it finish the mailbox it's currently on, or does it just end with the last message it's processed?

micah

[Michael_Bilsbor]

Hi,

We enter the schedule period and queue up a request for each mailbox.
We log into each mailbox and come up with a list of items to archive(up to a maximum of 1000 per 'pass'(configurable)). So we archive say 50 items from one mailbox because that's all that needed to be done, say 900 from another. If we find that more than 1000 items are eligible then we'll archive the first 1000 and then queue another request to do another archiving pass against that mailbox.
That means that we ensure that we don't let one big mailbox hog all of the processing time. We process all those initial mailbox processing requeusts and then go back and process those added requests from the larger mailboxes.

So to your first question we'll carry on processing the mailbox doing multiple scans until there is no more to do from that mailbox but once we've scanned all the mailboxes and archived what's necessary we don't scan them again during the schedule(after all there is no point)

For your second question it depends whether we've already processed that mailbox or not already by the time they are enabled.

For your last question If when we come to the end of the schedule we're still processing a mailbox then we'll finish that request. Nothing really to do with it being huge or not.

[TonySterling]

looks like Dodo beat me to it! ;)

If you are age based the task will continue to make multiple passes through the enabled mailboxes looking for archived items. By default, it grabs a 1000 items per pass(for EV6 and above).

If you are quota based the task makes one pass through the mailboxes. This is due to the fact it makes its initial calculation at the beginning of the archive run. Quota based can be precarious, and will most likely need refining to get right.

I always recommend age based to get the mailbox to where you want it. Then, if you have to, use quota to maintain it.

Are you asking if you enable a mailbox whilest the archive run is occurring? To be honest, it never occurred to me to try, but I would speculate it would not be processed till the next archive run, because it would need to be synchronized to let EV know it is enabled. --pure speculation ;)Message was edited by:
A WS

[Micah_Wyenn_2]

Awesome guys, thanks! That helps a lot.

micah