Hi All,
Just moved into a new job and got my first task: figure out if it's SQL Server that's causing users to wait a typical 45-60 seconds for GUI queries to return simple search results. The DB is less than 50 MB, there's less than 12 tables, all contain fewer than 100,000 records total. Users insist this extremely slow behavior is 24-7 and seems to have started after the latest version rollout. A query and stored procedure trace for 6 peak hours shows not a single query duration longer than 2 seconds. CPU, reads, and writes are all near ridiculous lows. DMV's suggest the same when checking total previous resource usage counts. From what the other DBA and I can see, this server is under no strain at all at any point during the last few days.
I don't want to throw my hands up and say, "it must be the network!" because I frankly don't know networks. I don't want to dodge or inappropriately push blame to another department, I want to be a professional about this and completely eliminate the possibility that SQL Server is causing the users' delays. What else should I eliminate? I've exhausted my understanding here, don't know enough about waits to know if they could be helpful here. I don't want to take on network administration and setup as a new hobby, but I want to know everything SQL Server can do to ID the cause of, and maybe even help resolve the users problem.
Note: I don't know a thing about replication, but I'm told some of the queries I see in the trace are generated by the replication. They all execute with a duration less than 2 seconds, and they're all identical, but they execute sometimes every 2 minutes, sometimes more than 4 times per minute. Is a poor replication setup liable to cause some delays in user queries that somehow wouldn't otherwise show up in a trace? (I'm grasping here, I know.)
Thanks,
Eric