SQL Server performance.

[DrSql]

I have same db in Server A and Server B
If I execute a sp in Server A it takes 1 min same time on Server B 1 sec.
The only diff is Server A is clusterd and data and ldf files are SAN drive. Now I am trying how to trouble shoot that. The SAN team everything is fine at there end. Any suggestions?

Dr.Sql
Good Luck.

 

[hmckillop]

We notice the performance on our cluster is poor because the database file is in one single filegroup and datafile.
Due to large amount of data (>90GB) and multiple accesses.

So the recommendation in this case is to split the data onto seperate filegroups, as per all the manuals.

You can use perfcounters to diagnose your problem, also check the obvious things such as indexes.
A SQL comparison tool would be the most obvious thing (such as SQL compare), this will definitely rule out different dbs.
Also use QA to profile on both and see that a wierd query plan hasnt been stored on one of the dbs (then run dbcc freeproccache, drop cleanbuffers etc).

Just my thoughts

"I'm living so far beyond my income that we may almost be said to be living apart

 

[timscronin]

How is the SAN configured, just because it is a SAN doesn't mean the RAID is wrong? WHat is the RAID on server B, what kind of drives etc. File groups kind of become a wash with RAID

 

[vnad]

I had a similar problem. Our data on server A was on a SAN and Server B had it on just a local raid array. Server B would be a lot quicker than server A. The issue turned out to be that they stripped the SAN incorrectly. Meaning, the partion had multiple strips on the same disks. So instead of being stripped 1 time on 10 disks, we were stripped 5 times on 2 disks.

I would check to see how the SAN is configured and ask what type of raid system they have for the SAN. Also check to see how many disks your are spread across. Might also want to check if any of the disks on the SAN are 'hot'. Meaning some other server/application is taping the i/o on the disk that you share your disk partition.