Viewing 15 posts - 44,491 through 44,505 (of 49,571 total)
What's filling up the drive? The log files themselves (.ldf) or the log backups?
What else is on the drive?
September 16, 2008 at 3:00 am
The missing indexes DMVs are reset when the server starts. They can't be modified by a user.
One thing I would like to add is that the missing indexes DMVs should...
September 16, 2008 at 3:00 am
Enable traceflag 1222 (DBCC TRACEON (1222.-1))
With that on, a deadlock graph is written to the error log when a deadlock occurs. You can read through the graph to see what...
September 16, 2008 at 2:56 am
Because none of the strings you have there are reaching the 30/40 character limit for the varchar. The ISNULL returns for the first query a data type of VARCHAR(30) and...
September 16, 2008 at 2:55 am
Already asked in the appropriate Question of the Day thread
http://www.sqlservercentral.com/Forums/Topic565885-1370-1.aspx
Please don't post multiple threads with the same question. It wastes people's time, it breaks replies up and results in people...
September 16, 2008 at 2:45 am
The last 2 may help. Depends where the slow down is coming from.
Check your IO performance when the index defrags are running. I'd suggest the following counters:
Physical disk:avg sec/read
Physical disk:avg...
September 16, 2008 at 2:31 am
Get more drives.
Can you explain a bit more? What drives are filling up (and with what) and where are things configured?
September 16, 2008 at 2:30 am
If you print out the dynamic SQL before you exec it, what do you see?
I tested a section of your code out
DECLARE @Stamp varchar(25), @global_time int
SET @global_time = 25
set @stamp...
September 16, 2008 at 2:24 am
Your query, as written, cannot use index seeks. The computation in the where clause means that the only index usage possible is a scan. Also, all or your indexes are...
September 16, 2008 at 2:08 am
Depends. 😉 On the stripe size, what kind of writes are done, and a bunch of other things.
With a 4 disk RAID 5 if a write only affects one of...
September 15, 2008 at 3:19 pm
I would go for option 2, simply because RAID 5 has the slowest writes of any of the commonly used RAID levels
September 15, 2008 at 3:03 pm
Please don't shout (all caps)
Unless there's some form of auditing in place, there's no way to know when a table/index was last updated.
September 15, 2008 at 3:02 pm
Table schema and sample data please?
I suspect what you need is a NOT EXISTS rather than a !=. You're looking for patients where there are no matching rows in the...
September 15, 2008 at 1:28 pm
There's not really enough information to answer the first question. Could you post the schema of the two tables and some sample data please?
Try using replace instead of the substrings,...
September 15, 2008 at 1:17 pm
You'd need to set up a linked server (Books online is your friend)
Assuming you have a linked server called server2 and database2 is on that remote server, the query would...
September 15, 2008 at 12:13 pm
Viewing 15 posts - 44,491 through 44,505 (of 49,571 total)