Fairly soon, I’ve got to get on an upgrade a couple of virtual servers I run and so I’ve been taking a fresh look at backup scripts. So far, I’ve mainly been relying on backing them up as a whole and also having copies of most of the data on them elsewhere but I want to make sure I’ve got everything I need covered as I’m minded to do clean installs this time round (and then restore settings).

Step one was identifying the areas where key data lived and rolling it up in compressed ‘tarballs’. Those can then be stored somewhere safe (itself backed up) and I can then restore as required. Easy enough but then I spotted that one of the sections was larger and so took a while to complete.

Phase two was to only run the backup for each area if things had changed since the last one was done. My approach was to create a ‘flag’ file, which is updated on each run. If I find any files in a section of the file system that are newer than the flag file, I’ll update the backup of that chunk; otherwise, I can wait and test again next time.

Seems to be working well so far so on with planning for the server upgrades.