One of the problems with working in ‘IT support’ is that the phone can ring at any time.
Sure – we sign up for this when we take the job and there’s no point in complaining that a Sunday morning breakfast in a little diner cafe turns into a 10 minute dash home, 10 minutes frustration with VNC and Contivity and a 30 minute drive into the office just so I can start a 6 hour long marathon getting things up and running.
I don’t mind doing support during the night, at weekends, whenever. It’s what I signed up for when I got into support in the first place. These days I’m on call around one week in three, previous to this job it was 24×7, 364 days a year.
I do mind, however, when the things I’m being asked to support are not within my control. Why do people reboot UNIX servers on a weekly basis? Why do sysadmins still rely heavily on cron for starting services (especially when there are FLOSS alternatives like launchd? Why did this process not start up the way it’s meant to – who changed the documentation? What’s the point in setting up two redundant servers when you need both of them up and running or your environment is fscked?
Why? Why? Why? Why? Why?