Got myself a little downtime here. Not sure what’s going on but the system SSD threw IO errors and the LVM switched to read-only as a result. I’m kinda amazed that the system kept chugging along like that at all. Anyway, a long SMART self-test came back clean and the system reboots fine. Just to end up in read-only again after a few hours. Sadly disks are really very expensive currently so I grabbed an older SSD 840 EVO from the pile and set the system up again this morning. That’s also not a disk designed for 24/7 but that doesn’t really matter for my use-case here. Biggest headache was Debian renaming network interfaces *yet again* so I had to figure out what was going on when no link could be detected.
Data is fine. That’s all stored on thicc SAS disks powered by ZFS and has various backups going. Just the system disk isn’t. This is because it’s connected to another SATA controller that was designed for CD-ROM usage and can deal with only one device on it’s bus. This is mostly because the system was initially configured with a hardware raid controller by HP and getting that to boot is a pain in the ass – and also doesn’t play nice with ZFS. That thing is out of the equation though and set to pass-through for years but I still like to separate the concerns.
I was not kidding when I added this disclaimer to the page: +++ This is a backyard hosted website on 2nd hand hardware, which means it sometimes goes offline. Please hold the line. +++
So anyway, everything should be working again – here is to another 4 years or so 🤓
https://beko.famkos.net/2026/05/11/15631508/