Deploy McAfee Agent Server Task hungs Epo 5.10

[User25693301]

This task submitted at 11:13:22 remains with the status "Waiting" The task is executed if the server is restarted

[cdinet]

Please check kb83384 to see if it applies.  

[philrandal · Accepted solution]

Update 1 for ePO 5.10 fixed this for us.  Available from McAfee support.

[User25693301]

The kb83384 was applied
the problem keeps on

[tomphill]

I am having a similar issue and it seems kb83384 no longer exists... Can anyone tell me what that KB might have said?

Thanks

[cdinet]

KB83384 was for where the internal dbcleanup task wasn't updating its runtime any longer and not running.  You first want to validate whether the cleanup task is running at all or not, or even if it exists - if you haven't installed any updates yet in epo 5.10, the base install was missing that internal task.  So, you first want to run this:

Select * from OrionTaskQueueMT where taskdescription like '%DB Clean-up%'

It should return 2 results, one that runtime and enqueued time update every minute when you re-run the query, and one that updates every hour.  That second one is for policy cleanup, so that one is not relevant to this issue.  You need to validate that those times are updating.

If, however, it only returns one result that updates every hour, or none, then apply CU4.  If you have already, run it again with the repair option.  If that doesn't fix, then follow kb94114.

If the task exists but are not updating every minute, it may be due to a backlog in the datachannel work queue.  

Run this script - select count (*) from epoagenthandlerdatachannelwqmt

If that count is large, then you want to truncate that table.  See attached script.

Then, check your enqueued and runtimes again for the dbcleanup task to see if they start updating.  If they don't, this is the query that was in the now unpublished kb (because it shouldn't be relevant anymore, as that was an issue with daylight savings time and is fixed in later versions of epo).

stop epo services
update OrionTaskQueueMT set RunTime=GETUTCDATE() where taskdescription like '%dbclean%'
update OrionTaskQueueMT set RunTime=GETUTCDATE() where taskdescription like '%task.queue%'
start epo services

[tomphill]

Thank you for that. I do have both results for the first SQL query. I am on EPO 5.10 update 4.

I rebooted the server prior to doing this, and that seems to have cleared it up...so I will keep this handy for next time. Seems to happen once a month or so.

Out of curiosity, could this happen if the DB server was rebooted and the EPO server was not rebooted? Just wondering if they maybe patched our DB server and rebooted and my EPO server only partially connected back to the database (log ins and everything else seems to work, just not server tasks).

[cdinet]

Yes, that is possible since some of those tasks require a response for the status for completion.  

[tomphill]

Perfect. Thanks again for your help!

[cdinet]

Glad to assist!  You can also mark as solution whatever fixed it for you.  There are sometimes more than one solution and others can try those solutions when needed.