Day 5 included cleaning up some duplicated CCTV footage from my TrueNAS server, 22 terabytes of it.
I run both BlueIris (for 24/7 recording) and Frigate for AI categorized motion alerts but didn't realize I had turned on full recording on Frigate and that had started to eat up lots of space. Frigate was saving in 10 second files so there were tens of thousands of them to delete.
The second item was to identify some of the previously failed disks in my disk shelf that I have replaced with spares. I had removed them from the cluster but they are still plugged in so I'm still getting bad sector alerts from time to time.
Those disks are reported with their disk identifiers like
sdz but not their serial numbers so you have to run something like this:
hdparm -I /dev/sd? | grep -e 'Serial\ Number' -e '/sd'
The next step is obviously to physically remove them. Thankfully when installing I had labelled all of the drive caddies with the serials so that made the job much easier.