Viewing 15 posts - 23,191 through 23,205 (of 59,102 total)
From what I've seen, Grant Fritchey's recommendations for changing the default threshold of parallelism are much more effective than changing the whole server to use MAXDOP 1. It keeps...
--Jeff Moden
Change is inevitable... Change for the better is not.
July 17, 2014 at 8:16 pm
Alan.B (2/7/2014)
--Jeff Moden
Change is inevitable... Change for the better is not.
July 17, 2014 at 7:55 pm
fregatepllada (7/17/2014)
Thanks Luis, I hate to bring another dimension to performance test but I am using TABLE variable, not temporary table. 😉
Heh... let me guess... "small row count". 😀
--Jeff Moden
Change is inevitable... Change for the better is not.
July 17, 2014 at 7:41 pm
fregatepllada (7/17/2014)
Thanks Jeff - I did CROSSTAB implementation before SQL 2005.For my current project I stick with PIVOT (I have less than 10K rows). 😉
Why would you intentionally...
--Jeff Moden
Change is inevitable... Change for the better is not.
July 17, 2014 at 7:40 pm
fregatepllada (7/17/2014)
IFAIK PIVOT became part of built-in SQL Server functionality since SQL Server 2005....
--Jeff Moden
Change is inevitable... Change for the better is not.
July 17, 2014 at 5:58 pm
That works. It's almost 2:30AM so I've got to hit the rack so I can be up at 6. I'll give this a whirl tomorrow night after work....
--Jeff Moden
Change is inevitable... Change for the better is not.
July 17, 2014 at 12:29 am
@marg 14154,
Are the 7 digit values for DN record type unique for each file? I ask because I have an idea for this but need to know that to...
--Jeff Moden
Change is inevitable... Change for the better is not.
July 16, 2014 at 11:35 pm
It doesn't sound like multiple databases would be the thing to do. It sounds like one database with some properly normalized tables would do.
This also sounds like homework because...
--Jeff Moden
Change is inevitable... Change for the better is not.
July 16, 2014 at 11:05 pm
inevercheckthis2002 (7/16/2014)
--Jeff Moden
Change is inevitable... Change for the better is not.
July 16, 2014 at 11:03 pm
Not a solution but some advice... rCTEs are frequently slower than a While loop and almost always use gobs more logical reads. I would be so quick to replace...
--Jeff Moden
Change is inevitable... Change for the better is not.
July 16, 2014 at 10:52 pm
murnilim9 (7/16/2014)
Btw I performed rebuild indexes on certain databases ..maybe around 22 indexes need to be rebuilt and it failed and showed an error...
--Jeff Moden
Change is inevitable... Change for the better is not.
July 16, 2014 at 10:45 pm
jjf112 (7/14/2014)
--Jeff Moden
Change is inevitable... Change for the better is not.
July 16, 2014 at 12:39 am
Igor Micev (7/14/2014)
--Extend this index to make it coverable for the query:CREATE UNIQUE NONCLUSTERED INDEX [UIX_did_uid_sid] ON [Dim].[UserDevice]
(
[DeviceID] ASC,
[UserID] ASC,
[SubscriberID] ASC
)
INCLUDE(
Locale,[LastUsedDate]
)
--Create this index
CREATE UNIQUE NONCLUSTERED INDEX [UIX_NCSTransaction_IDs] ON [Ext].[NCSTransaction]
(
[SubscriberID],[UserID],[DeviceID]
)
INCLUDE ([Locale],location,[TransactionStartDateTime])
If...
--Jeff Moden
Change is inevitable... Change for the better is not.
July 16, 2014 at 12:37 am
gurbanov.1984 (7/15/2014)
--Jeff Moden
Change is inevitable... Change for the better is not.
July 16, 2014 at 12:00 am
I don't know for sure but when was the last time you rebuilt stats and are you using just sampled stats or full stats?
--Jeff Moden
Change is inevitable... Change for the better is not.
July 15, 2014 at 11:53 pm
Viewing 15 posts - 23,191 through 23,205 (of 59,102 total)