Logging was wearing out my SSD

Just a reminder for amateurs like myself to check on journald, my elderly SSD has been wearing faster than it should and I just realised a Jellyfin process has been crash dumping every second for the last year. Journald was logging about 1.6GB per hour as a result, or around 14TB of writes a year. Nearly a terabyte per day. Whoopsie

Edit:

Didn't notice jellyfin crashing because I switched to plex. 14TB isn't a lot in the scheme of things but my old Evo 960 was nearly at 4x expected TBW so this possibly pushed it over into failure/noticeable performance loss.

I used iotop to figure it out then viewed journald live with journalctl -f

EDIT 2:

So I looked back through the SMART logs and the SSD has been writing nearly a F*CKIN TERABYTE EACH DAY for at least the last six months. How you might ask. Well systemd-coredump is doing a ~250mb write for every crash, however I couldn't see it in iotop because it's too fast - the coredump process starts and exits too quickly to see most times at the default refresh rate. Because of the fast NVME write speed it takes place in much less than a second. If I set the iotop refresh manually to 0.1 seconds you start to see it. No wonder its been running warm! WHOOPSIE