[97 MPWR]

Background info:
Windows 2003 SP2
BES: 4.1.6.12
Domino 7.0.2

We've had two outages this week and the high ups are fuming. The monitoring we have in place from a Windows standpoint has not provided any alerts on either crash.

Blackberry support response to the first crash was...

It appears that we have seen this issue previously under SDR209561. After the exception is caught, the BES then will go into a hung state due to the fact that every subsequent call to the Notes APIs fails. This SDR is targeted for the BlackBerry Enterprise Server 4.1.7 code line.

We are moving in to our busiest time of the year and Blackberry service is critical. We are also in a lockdown and even if 4.1.7 was released, it would be very difficult to get this implemented until early January.

I realize server hangs are difficult to monitor, but was just looking for ideas regarding what some of you are currently doing. Thanks in advance for any suggestions.

[qc_metal]

Do you know if there are any Windows event log entries that occur during or prior to the time when the problem occured? Or perhaps an entry in one of the BES log files? If so, maybe you can configure your monitoring solution to watch for those problems that lead up to the issue...

[m4ilm4n]

Tough one - my guess is if BES is in a hung state there won't be anything in the Windows logs or on the Domino console to indicate an issue. What I've noticed before is that when a process hung on our Domino/BES server, one of the CPUs went to 100% - you may want to try a perfmon trap for high CPU.

[hookedOnBB]

You can monitor for messaging activity (messages sent/received) and determine that the BES is hung if there is no messaging. The monitoring has to take into account after hours when there is a valid reason for no activity.

[mahoward]

If you use GSX for domino monitoring they have a BB module that can ping a particular device every minute with a return receipt enabled, and if it doesn't get a receipt back it will alert.

[auwcl]

May want to consider BoxTone - automates log-level in BlackBerry for Domino environment to identify individual non-responsive users and groups of users.