I am receiving the message "[50020] (10/27 08:53:38):{0xD4} Some worker threads have been blocked for 2 health checks" in the debugging logs on the Blackberry Enterprise Server. It does not appear to be affecting the Blackberry's functionality, but I would like to resolve this anyway. Anyone have any ideas?
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations IamaSherpa on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
worker threads being blocked for health checks?
- Thread starter mb004091
- Start date
StevePruitt
IS-IT--Management
We are getting the same errors. After a while, a BES server reboot is required - otherwise some users stop receiving messages. We haven't found the cause or a resolution yet, just doing the reboots as needed.
Steve
Steve
- Thread starter
- #3
What it seemed to be was one person's email was being held up on the server, due to his location and lack of signal strength. Once that user returned to an area where the signal was strong enough to receive emails, we stopped receiving those messages in the logs. Of course, this could be purely coincidental, but we haven't had the errors since.
Matt
Matt
StevePruitt
IS-IT--Management
I really hope that's a coincidence, Matt. First off the server really should be able to handle that easily. Second, that'll be a common thing for us with 150+ users. My own desk is in a dead zone, in fact. I have to leave the room for my Blackberry to catch up on messages.
Please keep me posted. Our warnings went away after a reboot Sunday, they usually take a day or two to resume.
Steve
Please keep me posted. Our warnings went away after a reboot Sunday, they usually take a day or two to resume.
Steve
- Thread starter
- #5
In our instance, the user had 1000+ messages pending to his handheld. I don't know if there is a threshold he crossed or what. I have not been able to find anything anywhere regarding this message. We have not received those messages in the logs for about a month now. Our building tends to be a dead zone as well, although, we have just under 100 Blackberry users. If I find anything regarding this, I'll post it here.
Matt
Matt
StevePruitt
IS-IT--Management
We had a flood of these messages Monday, and it'd tapered off to almost none today. I am suspecting a connection, given your observations, to some related problems we had then. In particular, the trust with one domain was broken so a number of BB accounts had problems because the BES and Exchange servers couldn't talk with each other.
We commonly have some users with huge numbers of queued messages for a variety of reasons. Most often it's units being transferred and the server accounts aren't updated, but other things too.
At least I did manage to get our vendors to sit up and try to help our server engineers on this. A senior RIM engineer said this was a known problem with an earlier version of BES software, but not with the latest version that we're running. We'll see what develops from that. I'll let you know.
Steve
We commonly have some users with huge numbers of queued messages for a variety of reasons. Most often it's units being transferred and the server accounts aren't updated, but other things too.
At least I did manage to get our vendors to sit up and try to help our server engineers on this. A senior RIM engineer said this was a known problem with an earlier version of BES software, but not with the latest version that we're running. We'll see what develops from that. I'll let you know.
Steve
StevePruitt
IS-IT--Management
We found that we have a memory leak on our BES servers. RIM is investigating, it's apparently not a known problem.
Also, we learned that BES uses SQL databases on it's own server by default. We're switching to using databases on our regular SQL server to reduce the strain on the BES servers, at RIM's suggestion.
Steve
Also, we learned that BES uses SQL databases on it's own server by default. We're switching to using databases on our regular SQL server to reduce the strain on the BES servers, at RIM's suggestion.
Steve
StevePruitt
IS-IT--Management
BES SP1a is supposed to resolve the memory leaks. We are installing it this weekend. I'll report back on results.
Steve
Steve
- Thread starter
- #9
Great. Only, for our organization, it won't help. We're running BES 2.1 SP5. Hopefully, we can either upgrade to 3.6 (since Exchange 2003 is coming to us), or SP6 for BES 2.1 will fix it. That is, if they decide to even put out another service pack for 2.1. How did you change where to store the SQL database? I looked all over and couldn't find it. Is it a registry "hack"?
Matt
Matt
StevePruitt
IS-IT--Management
We're running BES 3.6, which may make a big difference in the SQL procedure. Our engineers asked RIM the same question, they replied Not a problem to move. Just change the BES MGMT console to point to the new SQL server and db. See page 89-90 of attached doc for details." They attached the BES 3.6 Administrators Guide (available from the BB website).
And we're still having the same threads blocked warning messages. Two reboots so far this weekend. <sigh>
Steve
And we're still having the same threads blocked warning messages. Two reboots so far this weekend. <sigh>
Steve
StevePruitt
IS-IT--Management
Some updates to my previous messages:
1) A RIM engineer wrote to us saying, "If you are only seeing an average of up to 6 non responsive threads per day, this is nothing to be too concerned about. You're going to see some non responsive threads in any busy MAPI environment. It's the amount of time they are non responsive for that you should be looking at. If it's anything more then 5 health checks then maybe we should take a look at the logs. Anything less then that is common. The only thing I would suggest would be to apply Service Pack 2 as soon as it is released (we're looking at around the end of this month at this point). This service pack will as well upgrade the current threading model the BES is using to communicate with Exchange."
2) We haven't had reports of service problems since installing SP1A, so we're going to stop rebooting so much and see what happens.
Steve
1) A RIM engineer wrote to us saying, "If you are only seeing an average of up to 6 non responsive threads per day, this is nothing to be too concerned about. You're going to see some non responsive threads in any busy MAPI environment. It's the amount of time they are non responsive for that you should be looking at. If it's anything more then 5 health checks then maybe we should take a look at the logs. Anything less then that is common. The only thing I would suggest would be to apply Service Pack 2 as soon as it is released (we're looking at around the end of this month at this point). This service pack will as well upgrade the current threading model the BES is using to communicate with Exchange."
2) We haven't had reports of service problems since installing SP1A, so we're going to stop rebooting so much and see what happens.
Steve
StevePruitt
IS-IT--Management
BES 3.6 SP2 cured this problem for us.
Steve
Steve
- Status