SQL Server MVP Deep Dives 2, Chapter 9
I can file this chapter with many of the others under, “Things I wish I had read earlier in my career”. A good part of that career was at a small company and we were able to gauge our database capacity by the number of clients we had with pretty good results. This worked well because as we added more clients we had capital to add more database resources. Until…we wanted to add a very large number of clients all at once. People came to me for estimates on what we would need to handle the sudden jump and I could give pretty good estimates for disk space, but I had not been tracking our capacity for other things like memory, CPU and I/O.I knew the infrastructure we had would not be up to the task, but I didn’t have the information to back it up. Greg Larsen’s chapter in Deep Dives not only explains why you need to have these metrics, but he provides examples on which specific counters you should start looking at and how to gather them.
You won’t believe how quickly management will authorize the purchasing of a new machine when you produce a nice colorful graph showing how the various performance metrics are getting close to reaching your machine’s theoretical hardware limits. – Chapter 9, SQL Server MVP Deep Dives 2 Continue reading