Hi

Having a few troubles with a new BES box(latest version) that we recently put into Sydney. Its sat next to the Exchange boxes over there so there aren't any problems with that.

At the moment, were getting quite lengthy periods where no mail gets delivered to the Blackberry's.

Looking through the logs (MAGT) we see quite a few of these messages.

[40720] (12/04 13:53:38.189):{0x10B4} MAPI call failed. Error 'Network problems are preventing connection to the Microsoft Exchange Server computer.', LowLevelError 1726, Component 'Microsoft Exchange Server Information Store', Context 1300

Now there's been no problems with Exchange and connections to it.

At the moment the BES admin account is on the UK Exchange boxes. Now with a ping time of ~300ms to the UK i'm guessing the BES account is timing out and therefore losing connection. Therefore it can't rescan the mailboxes and can't tell that theres mail to be delivered.

I am planning on creating a new BES account on the Exchange boxes over there which i'm hoping should eliminate this problem but was just wondering if anyone else had had similar problems or could point anything obvious out?

Cheers,

Get the BESAdmin mailbox moved ASAP; MAPI is dying.

and dying hard core...

Max ping time for MAPI based on Microsoft recommendations is 30 ms.

Right, i've created the new account thats now local to the BES & Exchange. Swapped over services to use it and rebooted.

It seemed to work initially but then i saw these in the MAGT logs and since no mails been recieved.

[30054] (12/14 14:09:14.316):{0xA48} Starting Message System
[30055] (12/14 14:09:14.316):{0xA48} Using MAPI profile BlackBerryServer
[40206] (12/14 14:09:14.332):{0xA48} MailboxManager::SubsystemInitialize - Using MAPI profile 'BlackBerryServer'
[50000] (12/14 14:09:14.347):{0xBF8} Controller: This BES Agent is under control of BlackBerry Agent Controller
[20137] (12/14 14:10:51.813):{0xA48} MailboxManager::SubsystemInitialize - g_pSession->OpenMsgStore (0x8004011d)
[10277] (12/14 14:10:51.813):{0xA48} BlackBerry Messaging Agent BESSERVER Agent 1 failed to start. Error code 5305
[50106] (12/14 14:10:51.813):{0xA48} Stopping BlackBerry Mailbox Agent 1 for Server BESSERVER
[40507] (12/14 14:10:52.344):{0xBF8} ExtUDPLogThread: Code on closing = 0
[30000] (12/14 14:10:52.344):{0xBF8} [DIAG] EVENT=Unregister_thread, THREADID=0xBF8, THREADNAME="ExtUDPLogThread"
[30065] (12/14 14:10:52.657):{0xA48} BlackBerry Mailbox Agent 1 for Server BESSERVER shutdown complete

Thats roughly when users stopped receiving mails by the looks of things.

I'm sure everythings been created correctly for the account, and have made sure that I changed the mapi profile on the BES to point to the new account so i'm at a little bit of a loss of what to do next?

Any pointers of which logs could be giving some useful info?

Cheers

further to this it seems as though restarting services kicks into gear.

I'll check back later though as i'll presume thats its going to start failing again.

CHeers

I'm not surprised, but once it starts processing stuff over a length of time it'll slow down again.

I think that's the recomendation under normal conditions, not the maximum.

<35ms Ideal
<50ms Acceptable
<75ms Warning
>100ms Not possible

ping times between the bes and exch are now ~1ms. We created a new admin account on the local exchange box as previously it was on the uk exchange with a ping time of 300ms which was obviously killing it.

All reg entries of the MAPI profiles that the bes server is using are pointing at the new account on the local exchange. It did seem to die overnight several times though.

Again, i'll continue to monitor and post back later

Cheers for the help so far

[40720] (12/14 23:59:40.934):{0x96C} MAPI call failed. Error 'Network problems are preventing connection to the Microsoft Exchange Server computer.', LowLevelError 1722, Component 'Microsoft Exchange Server Information Store', Context 1300
[40000] (12/14 23:59:40.934):{0x96C} {adent@blah.com} Processed MULTI_STATS from Dispatcher, Tag=2363
[40720] (12/14 23:59:40.934):{0x96C} MAPI call failed. Error 'Network problems are preventing connection to the Microsoft Exchange Server computer.', LowLevelError 1722, Component 'Microsoft Exchange Server Information Store', Context 1300
[20301] (12/14 23:59:40.934):{0x96C} {adent@blah.com} Unable to save configuration settings or statistics
[40688] (12/14 23:59:40.934):{0xF50} {adent@blah.com} Starting message deleted rescan

These just popped up and i'm guessing mails been stopped.

Now this BES is running in a VM on the FS that we have over there. I can no longer ping from the BES to EXC but from FS to EXC is fine.

I'm starting to wonder if theres an issue with the VM as the EXC is always reachable from elsewhere.

Anyone had similar experiences?

bumpage.

Pretty sure this is more related to a VMware issue and the fact that that keeps losing connectivity rather than the BES Setup.

So.. anyone had experiences with BES in a VM before?

Cheers

I suppose it could be VMWare problem ... but I've used both VMWare and MS Virtual Server with BES and have never seen these problems isolated to a VM problem.

I have 2 BES servers in VMWare and never had these issues.....

Do you have T-Support?

What version of VMWare are you using? GSX or ESX? The errors that you are getting are showing that there is a network connection issue from the BES to Exchange. Is the version of MAPI32.dll matching on the BES and Exchange machines? If the Exchange is not pingable from the BES that would indicate an issue if it should be pingable. Not to sound stupid but what is FS? Front-end Serer?

cheers for the replys guys.

Were using the free version of VMWare Server at the moment. Unfortunately we don't have a T-Support contract but I do believe that since we've recently bought a couple of new copies of BES we may be able to get some support.

I've checked mapi versions and they do match.

Oh and by FS i'd meant FileServer

Its a weird issue... pinging from the FS (on which the vm is sat) to Exchange is fine. Ping at the same time from the BES VM times out for a couple minutes at a time. Can't tell at the mo how often it happens but it loses the connection for quite a while then obviously has to rescan all the mailboxes when it comes back up, resulting in the mail arriving in batches.

One last question for now though. We brought the software through our UK phone supplier and were obviously given a UK SRP to connect to. Would it be best to contact RIM to see if its possible to switch to one nearer Sydney?

Cheers again

Just a few last questions as unfortunately i'm still having problems with this install.

Starting over from scratch... what needs to be backed up before hand? There are only 4 users running on this BES at the mo so isn't really a problem to get them to re-activate but would be more ideal if they didn't need to.

Also could the fact that the box is located in Sydney and its using an SRP address of srp.uk.blackberry.net be a problem?

Document your license information; don't bother with anything else (unless you have IT Policy settings you want to keep track of) ... just wipe it all.

It could be causing problems ... but has this server always been using srp.uk.blackberry.net? As long as it has, and has solid connectivity (which I'm guessing it doesn't) it shouldn't have issues. You should call T-Support and let them know you want to use your SRP against a different SRP host and they'll walk you through the process.