Following the
installation of SP1, I have had a problem where both of my DCAM servers are
reporting "Incoming communications queue is full" and event ID 21268.
I have checked out the various KB articles on this and carried out the
following:
- Increased temporary
storage space on DCAMs to 100Mb
- Deleted *.PQF files
The DB has over 40% of free
space. The SQL agent job "Check Integrity" has been running for 2 -3 hours
(without any errors) - Can these jobs take a large amount of time to run
(our DB size is 5.1Gb) and could this be the cause and earlier I cancelled a
manual DBCC CHECKDB as it had taken hours with no result?
The agents have the messages "No connection could be made because the target
machine actively refused it" Event ID 21249 and although these are followed
by 21250 (contact resumed) we have now not received any alerts for the last
5 hours.
When the above has been done, the alerts start again for approx 5/10 minutes
and then stop again although the heartbeats are still coming through.
The dedicated SQL Server also reports Event ID 25107 "Failed to prepare
event for insertion into the database" and "lookup with insert returned no
data".
I have run the report for "Most Common Events" and it reports event counts I
would expect but the sample text shows as "Unable to find Security source".
Is this indicative of a SQL problem?
Contributed
By: Baelson Duque [MSFT]
There may be various reasons.
It might be that your DCAM is unable to insert data fast enough into the DB
so the DB maybe the bottleneck. Typically though, if you DB Server has been
sized appropriately, check if there are any collection rules that may be
enabled that are unnecessarily collecting a lot of events or Perf counters.
If you find a lot of these, start disabling some. This should lower the
amount of data being collected.
|
