Viewing 15 posts - 61 through 75 (of 275 total)
Try the following link...you may have to search for follow-up articles as there were some revisions/bug fixes to this code.
http://www.sqlteam.com/item.asp?ItemID=2955
March 1, 2007 at 1:29 am
to LauraS
"I suppose it could take that long depending on your hardware, how busy your db is and if there are any indexes involved"
If you can answer some of the...
February 26, 2007 at 6:46 am
"one more question why xp_cmdshell is recommended to be disabled?"
Because it runs under the security priviliges of the SQL Account....and can open a security hole on your server. Remember in...
February 7, 2007 at 2:22 am
Only thing I can think of would be to write triggers for ALL tables for all UPDATE/INSERT/DELETE statements to disallow same when the user comes from a specific IP/HOSTNAME.
February 1, 2007 at 2:19 am
options...
1. re-design the SP to guarantee to run within 1 minute.
2. re-structure your hardware to guarantee to run within 1 minute.
3. re-design your SP/Application to live with this SP not-running every...
January 31, 2007 at 3:36 am
Many thanks for the (re)ward. I was an early adopter of PromptSQL (@ 20$ it was worth a try), but my usage had vastly diminished before the Red-Gate...
January 26, 2007 at 8:32 am
I think this is related to using NT Authentication instead of SQL Authentication when signing into SQL. Experimentation on a TEST box would clarify same.
January 25, 2007 at 2:20 am
Personally...I'd also keep the PDF's out of SQL, and just store references to their actual LAN location. Putting them in SQL generates a large image management issue into SQL (backup/restore/read...
January 25, 2007 at 2:15 am
The issue of CACHE affecting the runtime's + physical read etc also needs to be considered. True testing would involve using the DBCC DROPCLEANBUFFERS + DBCC FREEPROCCACHE commands between runs.
I'd...
January 24, 2007 at 6:52 am
Don't know.....but if you can locate some links on SQLTeam.com....try asking the people who provided the answers. They have more direct experience of the problem and the solution.
Also....have you tried...
January 24, 2007 at 6:42 am
Have you recompiled the PROC.
Also looking into getting the actual execution plan.
And finally, you may be experiencing "Parameter Sniffing"...if you do a search of this site (and SQLTeam.com) you will...
January 22, 2007 at 2:47 am
"...nvarchar so Clustered Index is not recommended"
Whether or not the col is nvarchar or not is irrelevent for the Clustered Index...it's the cardinality of the column that affects is suitability...ie...
November 27, 2006 at 7:30 am
you have a lot of clustered-index scans....which I think are worse than tablescans....you want index seeks...for real performance. only way to do that is to rework the wildcards....
using the likes...
November 24, 2006 at 10:02 am
1. have a look at the execution plan, and try to see why indices aren't being fully used.
2. your LIKE statements....won't use indices...especially with leading %
3. qualify your tables with...
November 24, 2006 at 4:09 am
Don't know about that....but a workaround would be to have the jobs log their progress (finishing status) outside of SQL...and give your user access to that.
November 8, 2006 at 2:29 am
Viewing 15 posts - 61 through 75 (of 275 total)