SQL server 2008 R2 sizing

I have an application that end-users have reported to behave very slowly.
The application is on a seperate Win2008 R2 server, and uses a DB on a 2-node SQL 2008 R2 Standard cluster (on Win2008 R2 Enterprise).

I suspect that the SQL cluster is not running optimally. I have not started to log performance yet, as it is a little complicated (many people involved).

The SQL cluster is running 29 databases in one instance, some databases are around 130 GB in size, and the total size of DB's is 600 GB. The cluster is running SQL 2008 R2 Standard, with 64 GB RAM. I know that Standard edition of SQL only support 64 GB RAM, could this be an issue for this workload?

I am planning to migrate the disks (SAN, with RAID 0+1 on midrange 10 000 RPM SAS disks) to either better 15 000 RPM FC disks, or SSD disks.

But I need some  first impression from you guys.
My initial thought are that 64 GB RAM is way to low for the workload. I don't know of other applications with DB's on the same cluster have performance issues, but I suspect so. The instance is capped to use 60 GB of RAM, leaving 4 GB available.

So what do you think? Might be a problem?
LVL 9
xcomiiiAsked:
Who is Participating?

[Webinar] Streamline your web hosting managementRegister Today

x
 
Scott PletcherConnect With a Mentor Senior DBACommented:
You can run the code to see if any "low memory" conditions are detected.  If not, you probably don't have any significant mem pressure at the moment to worry about.

Have you looked at the Top Avg and Total I/O queries?  Sometimes they'll be a "bad" query or two that will really slam a server's performance.  Nowadays issues tend to be RAM or I/O rather than CPU (although that's not 100% guaranteed, of course, esp. if you do lots of xml in SQL).
0
 
Vitor MontalvãoMSSQL Senior EngineerCommented:
The easiest thing is to think that's an hardware issue but more than 90% of the times it isn't.
I recommend you to launch a SQL Profiler to check for blocking processes and long running queries. Capture those queries and analyze them one by one. You might need to create indexes to improve the performance of those queries.
0
All Courses

From novice to tech pro — start learning today.