[Legros]

Hello,

We are using BES 4.0 for Groupwise (6.5). We have 70 users... In the last month we had some problem with the server ( no more message to handheld, no attachment, etc). Each time the solution was to restart the service or the server. Each time, i'm amazed about the time it takes before the server finally comes back and start to deliver some message (around 2 hours...). Seems too long to be correct since we only have 70 users... ??

Thanks.

[bfrye]

Are you using MSDE or SQL? We notice the same thing with MSDE (upgrade procedure in the works)

[Legros]

Hello,

we also are on MSDE... We will check the opportunity to change for SQL. Were did you get the info that switching frorm MSDE to SQL gonna correct the problem?

tks.

[bfrye]

Within RIM tech support calls.

We're supposed to upgrade soon as SP1 is release and I'll let you know what the results are. We're more users than you though, and have noticed it can take 2-3hours for email to get flowing reliably again on reboots or controller restarts.

[udontknowjack]

RIMs pat answer to me on a reboot is "give it a couple of hours to scan the mailboxes before it starts sending" I have hardly ever had it take that long but a couple of times, It did take an hour or longer for some reason. Not sure why, on those occasions, but keep in mind, when you do a restart, the BES does have to do a rescan of all mailboxes and if you have a large pending count or large mailboxes to scan, it could take a while, especially if you don't have much ram or CPU on your server.

[bfrye]

That's our main problem. We have no service inside the building during the day, so our pending messages will easily reach 5,000, 7,500, sometimes even 10,000 during the day

[Snyder81]

I read a while back that MSDE only allows 8(?) concurrent database connections. That would explain why a reboot takes forever if it has to poll all the mailboxes. We moved to SQL Server (using a different box) and the reboots aren't too bad. Mail seems to start flowing within a few minutes.

However, there was one time when I rebuilt the 4.0 production server overnight(Had to wipe out MSDE database since migration to SQL initially failed) and it took over 12 hours before all the users were getting mail again!

[Somnambulant]

After you restart the BES, you get into the realm of SDR53831.

Here is RIM's description of that SDR:

"If the BlackBerry Messaging Agent is restarted and there are a large number of over-the-air folder management (OTAFM) commands in the WorkQueue table of the configuration database with a status of 0, users experience a delay, of up to a few hours, before receiving messages on the handheld."

That's from the SP1 release notes:

http://www.blackberry.com/knowledgec...46092&vernum=0

After you restart your server, you can look in the MAGT (Mailbox Agent) log and you will see that OTAFM packets are being successfully delivered to the handheld. Meanwhile, the pending count goes through the roof. Once all of these OTAFM items are processed, mailflow returns to normal.

[Legros]

Switching to SQL and updating to SP1 did solve this issue and numerous other...

[bfrye]

If you are having a really hard time getting it to process those queued messages, there's a way to delete them. From your SQL/MSDE server command line you can type

C:>  osql -E
1> use BESMgmt
2> delete from WorkQueue where ObjectDefnID = 7000
3> go
xxxxxxxx rows affected