Habitual timekeeping

Early this Thursday morning I am to help out with a scheduled upgrade to a local production server. The server suffers a problem such that a user action can bring it down. I generated a reasonable hypothesis as to the cause but I have not been able to reproduce the severity of the server crash in my own testing. During the maintenance window we can try reproducing the crash on the very system that suffered it last. If it happens again then we will take what relevant readings we can before doing the planned upgrade. The upgrade plan includes a configuration change which should help if I am correct about the cause; we will then again try reproducing the crash before the maintenance window ends.

I caught myself thinking that I probably charge this production server maintenance to a different cost code. We moved to Scotland back in 2012. For over a decade previously in the US I was required to track my work time in fine detail. It has now been some time since I saw a timesheet but it appears to be difficult to wholly shake old habits.