May 29, 2009 at 8:16 am
I have been using SQLIO for some time to look at the raw performance of disk storage. I have a suite of tests that I run to look at sequential\random read\write performance for block sizes of 8K, 32K, 64K, 128K, and 256K. I have found it very useful for identifying serious IO problems, but I would like to be able to use it to predict real world performance.
I was wondering if there are any guidelines for the different types of IO that SQL Server performs under real world conditions so that I can use the different SQLIO results to estimate how SQL Server will do with specific tasks. For example, it appears that the performance that I see on an SQLIO 64K sequential write test corresponds closely with what I see when running a database restore, and the performance from an SQLIO 64K sequential read test corresponds closely with performance of DBCC CHECKDB and with a query that scans a large table.
May 29, 2009 at 9:00 am
I'm with you on this one.
Viewing 2 posts - 1 through 1 (of 1 total)
You must be logged in to reply to this topic. Login to reply