SQL Server Management Through Monitoring

How do you manage your SQL Server instances? No, not connect to an instance to push buttons and pull levers. That’s not what I am asking about here. What I want to know is, how do you effectively administer and manage your entire environment? Are you in a reactive mode, where you wait for someone to come to you with an issue for you to research? Or have you turned that corner, and are now able to be more proactive when administering and managing your enterprise?

Last month I was in New York City (and parts of New Jersey) to meet with clients. I spent a lot of time using public transportation to get around the city (and parts of New Jersey). I was impressed with my colleague David Waugh (no blog | no twitter | and he calls me a loser?) and his ability to navigate the public transit system. I mean, how the hell did he know that we could take a water taxi back from New Jersey and catch a town car to get us into midtown for a meeting in the next 20 minutes?

And while in that town car I started thinking about taxis. And their mechanics. And how most of those mechanics have but one mission to achieve each and every day: keep the cabs running. So everything they do is based upon keeping that fleet moving. While I am certain they perform preventative maintenance, I am willing to bet much of their work is more of the reactive mode. Here, fix my flat tire. Here, my cab is overheated. Here, the brakes are shot. Here, there are bullet holes in the trunk. Those mechanics spend more time fixing problems as they arise than they do being proactive.

Now, let’s consider another mechanic. How about one for an F1 racing team? What’s that? You don’t know what F1 racing is? OK, then, lets take out our teeth and our banjos and talk NASCAR and stock car racing. I’ll make it even easier for you: F1 is where Juan Pablo Montoya used to race. Is that better? Anyway, they have mechanics, right? But how much of their job is reactive, and how much is proactive? Just keeping their car running is not enough, they need to keep it running in peak condition, especially on race day. Sure, the F1 mechanics are reactive to some degree, but many of their tasks are done for proactive reasons as they always need to stay one step ahead of the competition.

Now, which mechanic are you? Do you find yourself more in a reactive mode? Or are you able to be more proactive in your daily tasks?

And that’s the difference between those DBAs that manage their instances and those that monitor them. If you tend to manage your instances then you are more than likely going to be reactive in nature. If you are actively monitoring your instances then you are going to find more opportunities to be proactive. And the end result? Either your enterprise is functioning about as well as having Latka as your DBA or you have yourself a nice little F1 garage where you sip champagne and eat strawberries for lunch with your tea.

Once you start using tools to monitor your environment (shameless plug, I work for Confio Software, makers of Ignite, which you can download for free), you have the ability to perform a lot of preventative maintenance, and in turn you have the opportunity to keep your environment operation at peak efficiency. Oh, and chances are you can keep your phone from ringing as well. Instead of people calling you to complain that something is broken, they don’t need to call you at all because you’ve already fixed things before they became problems.

And when you get to that point, feel free to call yourself a senior DBA, you will have earned it by then.