[hurricaneguy]

We have several users that are not getting messages. They are on an Exchange server that is geographically separated from the BES. After doing some research I found the following post
Manually restarting a single Messaging Agent - BlackBerryForums.com : Your Number One BlackBerry Community

It describes the process of restarting the agent that has the hung threads. After doing this more users started to get emails but not all.

We have identified some problems. The major one is latency. The internet connection is being pounded at parts of the day and causing high ping times. We have tried to do the agent restart and even a whole reboot of the BES at some times when we have good ping times. But only some of the users get emails.

What we cant under stand is why only some of the users on Agent 1 are affected. We have another Exchange server that is local to the BES. All of the users on that are on Agent 2. None of the users on Agent 2 have problems.

So what would cause some users on one agent to have blocked threads and not others?

We have a BES 4.0.2 and Two Exchange 2000 servers, one local one remote.

[hurricaneguy]

Here are some of the logs we get.
From CTRL log
[30000] (06/10 14:50:19):{0x754} Performing system health check (BlackBerry Controller Version 4.0.2.8)
[30000] (06/10 14:50:55):{0x774} 'XXX-BLACKBERRY' agent 1: hung threads detected. WaitCount = 1
[30000] (06/10 15:00:19):{0x754} Performing system health check (BlackBerry Controller Version 4.0.2.8)
[30000] (06/10 15:00:55):{0x774} 'XXX-BLACKBERRY' agent 1: hung threads detected. WaitCount = 2
[30000] (06/10 15:10:19):{0x754} Performing system health check (BlackBerry Controller Version 4.0.2.8)
[30000] (06/10 15:10:55):{0x774} 'XXX-BLACKBERRY' agent 1: hung threads detected. WaitCount = 3


[30181] (06/10 15:10:54):{0x7E0} Performing system health check (BlackBerry Mailbox Agent 1 - BESX Version 4.0.2.8)
[30038] (06/10 15:10:54):{0x7E0} Thread: *** No Response *** Thread Id=0xCA4, Handle=0x7C0, WaitCount=3, RELOAD_FOLDERS event for User: aaa@yyy.com on Server: HFL-FTLD
[30038] (06/10 15:10:54):{0x7E0} Thread: *** No Response *** Thread Id=0xF50, Handle=0x874, WaitCount=3, RELOAD_FOLDERS event for User: bbb@yyy.com on Server: HFL-FTLD
[30038] (06/10 15:10:54):{0x7E0} Thread: *** No Response *** Thread Id=0xF54, Handle=0x888, WaitCount=3, RELOAD_FOLDERS event for User: ccc@yyy.com on Server: HFL-FTLD
[30038] (06/10 15:10:54):{0x7E0} Thread: *** No Response *** Thread Id=0xF58, Handle=0x898, WaitCount=3, RELOAD_FOLDERS event for User: ddd@yyy.com on Server: HFL-FTLD
[30038] (06/10 15:10:54):{0x7E0} Thread: *** No Response *** Thread Id=0xF5C, Handle=0x8A8, WaitCount=3, RELOAD_FOLDERS event for User: eee@yyy.com on Server: HFL-FTLD
[30038] (06/10 15:10:54):{0x7E0} Thread: *** No Response *** Thread Id=0xF78, Handle=0x918, WaitCount=3, RELOAD_FOLDERS event for User: fff@yyy.com on Server: HFL-FTLD
[20226] (06/10 15:10:54):{0x7E0} At least one worker thread seems to be blocked (3)

[hdawg]

Within the BlackBerry Manager check to see which messaging agent users are on. You can pick it from the list of options in there.

I haven't used BES 4.0.2 in a while ... I'd highly recommend updating to the latest 4.0 version (4.0.7 IIRC) ... however it is possible that some users have larger mailboxes / more items in the mailbox or the server just happens to be poorly processing their mailboxes.

You could try putting them on static agents ... but really the only proper solution is to move their mailbox or add a BES locally.

[MaxSmoke]

We ran into a similar situation here. We’ve always had a latency issue between BES and our Exchange servers in EU. While delivery was slow during peak times everything worked. This became a problem with two recent projects. First we started consolidating our mailbox servers in EU from 7 sites running Exchange 2003 to a single EU site running Exchange 2007. To support this we installed Cisco WAAS devices with MAPI compression turned on. Unfortunatly, these devices were undersized and running older code, 4.1.3.
During peak times the devices became overloaded and went into bypass mode. This increased the latency between my BES and the EU mailbox server. The WAAS device would then allow the IP keep alive to expire resulting in hung threads. There’s an unpublished RIM KB article in the works addressing keep alive settings on load balancers. We end up turning off MAPI compression until we can get the latest WAAS OS tested and deployed.
This has allowed mail delivery to continue though slightly delayed. However, we were unable to complete enterprise activations for staff on the EU mailbox server. To work around this issue we take these step take place after adding the new account and before setting the activation key

1. Double click the new account.
In the properties menu select PIM sync
Set Wireless Synchronization to False
In the properties menu select Advanced
Set Enable Static mailbox agent to True
Set Mailbox ID to 209

2. Complete the Activation process.

3. Double click the new account.
In the properties menu select PIM sync
Set Wireless Synchronization to True
In the properties menu select Advanced
Set Enable Static mailbox agent to False
Mailbox ID will now be grayed out.

4. On the device main screen click Messages.
The Messages screen appears.
Click the track wheel
click Reconcile Now.

The final solution will be to put a second BES in Europe and add it to the BES server pool.