We recently
upgraded to SP1 and all of a sudden our Database went up from a modest
(although large) 9 GB to a whopping 42 GB over a 2 week period.
I have reduced the amount of data from a previous 40-day period to a 2 week
period and Grooming is working properly but the Database stays around the 40
GB mark.
I think there may be other data that causes this huge size. How on earth do
I check which table is taking up all this space?
From the microsoft.public.mom
newsgroup
The first thing to do is check the report called 'Most Common Events'. I
suspect that you are collecting a bunch of events that you probably don't
want. If this is the case, you will see a high percentage of events of one
or two types. You obviously want to disable the collection rules for these
events if this is the issue.
The tables that will typically take a large amount of space in MOM are 'Events'
and 'EventParams'.
The 'SampledNumericData'
can have several million records, but the size of the data isn't usually
very big since it's just a couple of GUIDs and some integers. The event
tables get large because of varchars and such.
The best way to clean things up if you have too many events is to do a
temporary modification to the grooming jobs. These jobs will run every hour
between 11 PM and 5 AM. By default, each execution will only remove 4,000
records. If you have several million to remove, then this will obviously
take a long time. You might want to temporarily increase this number and
possibly increase the time window for it to run. You can set the batch size
in Global Settings in MOM but need to set the execution times in SQL under
the SQL Agent. Make sure you have the log set to grow since additional log
space made be required for additional deletes.
|
