April 1, 2003 at 10:08 am
Hello all,
I have a large table (15M + rows). On a daily basis this table has about 40,000 updates and 60,000 inserts - part of an OLAP db. The clustered index (two ID columns for the primary key) has a fill factor of 99%. According to the BOL, I would think it should be lower due to the updates. Would you agree?
I'm also unsure as to the degree it should be lowered...I've read a couple of good posts on the topic but thought this case was a bit different given the update/insert nature of the table.
Thanks!
April 1, 2003 at 1:10 pm
I think that you should let the fill factor high if for example the updates are, for example, from string of 100 char lenght to 100000 char lenght. In cases like this, leaving the fill factor low could still cause more page splits. But the most important for the fill factor value are the inserts, and with such a high value of inserts I don't think that it is a very high value. You could also check SQL for page splits for example, to check that the value is right or not.
April 1, 2003 at 1:23 pm
Yes. I agree with racosta that we have to concentrate on INSERTS only. But I think we should lower it to minimise the work on the SQL Server on page cascading.
.
April 1, 2003 at 1:55 pm
From Performance Monitor Counters taken during our load of this table I see SQL Server Access Methods: Page Splits/sec anywhere from 30 with a spike up to 240 - mostly sitting in the 110s.
April 1, 2003 at 3:56 pm
Giving a thought to that aspect, I think we can get reduce the pages/sec by reducing the fill factor.
Any Thoughts!!!!
.
Viewing 5 posts - 1 through 4 (of 4 total)
You must be logged in to reply to this topic. Login to reply