September 22, 2003 at 1:06 pm
Okay everyone! I have a table configured as follows:
Index on Primary Key - NonClustered
Secondary index on Foreign Key
Using DBCC DBREINDEX I can pack those nice little pages with data all the way up to the happy fill factor's fill. Great! That Scan Density is looking good, but WAIT, a quick glean of the page density is telling me that index with the id of 0 (zero) is only filling the page with that valuable data only 13% of the way. The problem is not that I am superstitious about the number 13; although there is some minimal irony here. The problem is that I have all these buckets (I mean pages) that aren't being fully utilized and everytime I need to look in these buckets for the tool I need (I mean data) I have to pick up all these buckets and look inside.
What I want to do is pack my data pages with data and reindexing is not reclaiming the space on the data pages of the table. I know I could fix this by temporarily turning the primary key to be a clustered index. For now, this is not what I want to do.
Is there any other way to pack the data pages on a non-clustered table?
Thanks,
Jeff
"Keep Your Stick On the Ice" ..Red Green
September 22, 2003 at 2:31 pm
How many pages do you have? If you have less than one full universal extent (8 pages) I've seen where you can not fill it any more efficiently.
Ray Higdon MCSE, MCDBA, CCNA
------------
Ray Higdon MCSE, MCDBA, CCNA
September 22, 2003 at 2:37 pm
quote:
Is there any other way to pack the data pages on a non-clustered table?
No; the only workaround is to create and then drop a clustered index. This is one of the problems with using heaps.
--Jonathan
--Jonathan
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply