Things I learned the hard way in IT:

• If a user claims there is “no vital data” on a computer, never believe them. Even if they insist, even if they claim you are wasting their time. Image the fucking hard drive if you are upgrading. Just image it. Users are completely clueless about where their data resides, or even if there is data.

• If a department claims “no one uses this server, it can be decommissioned,” first power it down. If no one complains after a month, remove it from the server room (if physical) or the virtual environment (if virtual) but DON’T GET RID OF IT for at least two years. I’ve had servers in environments that are only used once a year (at tax season, for instance). Two years full retention of server image (not just data) is usually enough.

• If a telecom vendor tells you 30 days, expect 90+. If they tell you 90 days, count on 180+.

• If a programming group complains about performance and blames infrastructure, learn enough about their code and their databases to be able to read the code and to debug their programming and DB issues. It’s almost never an infrastructure problem, and anyway throwing more hardware at a programming or database issue gets very, very expensive very, very quickly. In some environments, I’ve understood more about what the code was doing at the machine level than the programmers themselves did. (This is not actually uncommon among really good systems administrators. SAs have to know nearly everything about everything — programmers only have to know how to code.)

• If you have a manager in IT with an MBA, chances are you are going to have a bad time.

• If a user claims they have rebooted their machine, also never believe them.