We recently deployed a major software release, and ever since then we having been having huge issues with blocking due to pageiolatch_sh waits. I don't think there is a problem with disk performance since it started happening right after the release, but I suppose it's possible.
My question is: are there any tools or articles out there that can help me focus my investigation? I've been finding a lot of "it could be caused by this, or this, or this", but nothing to actually help me pinpoint the source of the problem (for 2000). I've found a bunch of articles out there for 2005, but not 2000. The entire database is bad and needs to be rewritten, but the database is huge and I need to fix this yesterday. Rolling back the release is not an option.
Any help would be greatly appreciated!
Thanks in advance,
I think there is a disk problem. More specifically, it appears you have deployed something that does table scans. Maybe a few big ones, maybe a lot of little ones. Hard to say from this information. Run a profiler trace, and pick out queries/stored procedures that are racking up a bunch of reads. See if you can rewrite those, or help them limp along with an index.