Viewing 15 posts - 226 through 240 (of 279 total)
You can disable non-clustered indexes, which frees up space, reducing the amount of data that needs moved in the shrinkfile.
You'll almost certainly need to rebuild your indexes after...
October 23, 2011 at 8:59 pm
What do the logs say?
c:\program files\Microsoft Sql Server\100\Setup Bootstrap\Log\xxxxxx_xxxxx\Detail.txt
where "xxxxxx_xxxxx " corresponds to the timestamp of the failed install.
October 20, 2011 at 10:26 am
truncate deallocates the pages.
delete flags rows as deleted, ghost cleanup task runs periodically in the background, and removes the rows from the page. If the page is then...
October 20, 2011 at 10:21 am
Any reason why you can't swap in a empty partition for the one that needs split, split the partition, then copy data back into it?
You're going to have to eat...
October 18, 2011 at 1:49 pm
No, that setting only affects the buffer pool memory, not the memory used for the CLR. If you do enable lock pages in memory, (which I'm not recommending as...
October 18, 2011 at 1:36 pm
Aspet Golestanian Namagerdi (10/17/2011)
October 17, 2011 at 10:09 pm
Without looking at your model, i'm going to guess and say go with a site dimension vs an attrib of task dimension.
Also, depending on the number of fact rows, you...
October 16, 2011 at 8:24 pm
Personal preference. Cte's format better with sql prompt, and if I build the query incrementally, cte's are easier to use.
Honestly, the only queries I've written recently have either been...
October 15, 2011 at 7:16 pm
Lexa (10/15/2011)
We're planning on testing a new storage device in production. We're planning on putting Temp DB and a couple user DBs on the device. Since we use a...
October 15, 2011 at 6:40 pm
Aspet Golestanian Namagerdi (10/15/2011)
NETWORK_IO 18 ...
October 15, 2011 at 4:21 pm
One option is to use a data flow, and relapce crlf with lf in each field, and save it back out. Might be problematic in ssis though...
The other option...
October 15, 2011 at 4:06 pm
Best approach is to combine them into a sql dw first, then load from there into the cube. If you can't do that, then linked severs combined with...
October 15, 2011 at 3:52 pm
If "id" uniquely identifies a row, it is appropriate to use it as the key. set visible to false so users of the cube don't see it.
October 15, 2011 at 3:46 pm
noyb 30173 (10/15/2011)
About VLF's...frighteningly, I ran dbcc loginfo and it returned over 833,000 rows!Could this be contributing to recovery times and how do I get it cleaned up??
we've definately found...
October 15, 2011 at 3:19 pm
10 indexes.... design done by a developer with limited sql experience... It makes me wonder if all those indexes are needed? If you want to post the table...
October 14, 2011 at 11:02 pm
Viewing 15 posts - 226 through 240 (of 279 total)