Viewing 15 posts - 61 through 75 (of 82 total)
Good point ... I'll have that checked
Leon
June 8, 2008 at 8:18 am
I believe there are a few potential improvements to be achieved by optimizing index definitions ..... but the big question troubling everyone here is when the databases are the same,...
June 7, 2008 at 9:42 pm
HappyCat,
Here is the response I get from IBM for the query on the disk cache -
Leon,
Sorry for the delay I can give you cache hit rates but to...
June 6, 2008 at 11:26 am
Thanks HappyCat! That helps ... I am going to check this out!
Leon
June 1, 2008 at 2:56 pm
That is what I thought ... man but IBM is such a pain the butt .. they just discounted that as "insignificant difference" ..... they keep saying Disk on Production...
May 29, 2008 at 10:30 pm
Does it help if I said the Database from production was backed up and restored in Development and then the test was run from the same application server that connected...
May 29, 2008 at 5:29 pm
also, if you buy into the idea that it is being swapped out completely (difficult though since it stays at 6GB) ... would setting a min server memory limit help?
vice...
May 27, 2008 at 7:30 pm
Barry
Thanks!
1. I agree ... these are significant considerations in the cost-benefit of locating the indexes on a seperate disk
2. I am sure missing something ... wouldn't truncate table clear all...
May 10, 2008 at 5:44 pm
One of the findings IBM reported was that the Production server was connected to the SAN through a single attached switch .... (not sure if I am saying this right)...
May 8, 2008 at 6:29 pm
I am inclined to believe so too .. Is there any reason why the sql server will be responsible for this slowdown? I want to make sure we (SQL team)...
May 8, 2008 at 7:01 am
I guess I overlooked something .. the number of write operations per second .. they have increased from 8 (when this ran per schedule) to 20. does look like contention?...
May 7, 2008 at 9:11 pm
I dont understand something with these perfmon numbers here .....
During the actual production scheduled run, disk sec/write were around 27 ms for the database and the write throughput was 180...
May 7, 2008 at 8:57 pm
Barry
Thanks! Its a Fast T. I guess it is being used by other servers too. So you are thinking contention could potentially cause this. I will bring this up with...
May 7, 2008 at 7:22 pm
While I was of the same opinion initially, I get the feeling we tend to overlook a few other areas that should be investigated (being a SQL Server DBA) before...
April 23, 2008 at 3:30 pm
Viewing 15 posts - 61 through 75 (of 82 total)