Just finished a systems planning document requested by T, the boss of my boss. It concerned the acquistion of some expensive software, at least by the standards we are used to. I did some rough estimates of systems reliability under different assumptions, then added up the corresponding costs. For example, in one case we could reduce the expected downtime from roughly 34 hours to about 2 minutes. However, it would cost over $150,000 more. Is it worthwhile? Not for me say. The people who need the systems, and hence pay the bills, will make that call. I am providing options and pointing out the consequences.

I suspect, with the numbers I worked out, most of the managers here would decide that the more expensive option is not justified. Unfortunately I think that T, who is new to the organization, wanted me to come up with a justification for the purchase. Well, as the CSI folks say, I have to go where the evidence leads.

In any case, the exercise was kind of fun. I got to compute, and explain, a few combined probabilities (e.g. chance of multiple servers failing) and the consequences. Actually quite simple stuff, but I don’t see much of it around here.

Next task: Build a crystal ball. Officially this is known as capacity planning.