Hi

I have been experiencing server failures forcing me to do hard boots.

My syslog has no logging for 24 hours until 16:03 tonight. Then no logging for 90 minutes before failure.

Nov 8 16:36:38 Tower Parity Check Tuning: ... but suppressed as system notifications do not appear to be enabled

Nov 9 16:03:54 Tower root: /mnt/cache_download: 931.5 GiB (1000203837440 bytes) trimmed on /dev/sdg1

Nov 9 16:03:54 Tower root: /var/lib/docker/btrfs: 6.6 GiB (7104176128 bytes) trimmed on /dev/loop2[/btrfs]

Nov 9 16:03:55 Tower kernel: sd 9:0:5:0: [sdg] tag#435 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=DRIVER_OK cmd_age=1s

Nov 9 16:03:55 Tower kernel: sd 9:0:5:0: [sdg] tag#435 Sense Key : 0x5 [current]

Nov 9 16:03:55 Tower kernel: sd 9:0:5:0: [sdg] tag#435 ASC=0x21 ASCQ=0x0

Nov 9 16:03:55 Tower kernel: sd 9:0:5:0: [sdg] tag#435 CDB: opcode=0x42 42 00 00 00 00 00 00 00 18 00

Nov 9 16:03:55 Tower kernel: critical target error, dev sdg, sector 1951699269 op 0x3:(DISCARD) flags 0x0 phys_seg 1 prio class 0

Nov 9 16:06:52 Tower root: /mnt/cache_protected: 83.0 GiB (89162940416 bytes) trimmed on /dev/sde1, /dev/sdh1

Nov 9 17:40:39 Tower rc.rsyslogd: Syslog server daemon... Started.

Is this the reason my server is failing or should I be looking for something else?

Nov 9 16:03:55 Tower kernel: critical target error, dev sdg, sector 1951699269 op 0x3:(DISCARD) flags 0x0 phys_seg 1 prio class 0

9 hours ago, davper said:

Then no logging for 90 minutes before failure.

That typically suggests a hardware problem; you can post the diagnostics mostly to see the hardware used.

Diagnostics attached

tower-diagnostics-20251110-0923.zip

Nothing jumps out as having known issues, since you have multiple RAM sticks, try using the server with just one, if the same try with the other one, that will basically rule out bad RAM.