MammothJerk

haven't had any crashes since, mover has mover 3TB and parity check finished (with lots of errors 🥲) and the server has not crashed.

I'm not sure i would call the issue "Solved" since the issue still persists but has a "workaround" for some setups. as tamet83 said, is it possible to add a delay on startup to apcupsd to check if that indeed fixes the problem? I tried turning off the daemon before a shutdown, and then turning it back on after boot. This did not work and i still got the garbled information. My UPS came with the rj45-db9 cable as well, plugging that in and setting it up correctly also worked for me. Though i had to reverse the master/slave configuration since my main server doesn't have a serial port available.

I also had this exact same issue but following the steps from @Gnomuz fixed it for me as well. It would also sometimes work for me if i power-cycled the UPS and then started the unraid server but as it was unreliable i was just using USB. Hopefully the issue can be fixed, not sure if it's unraid or apcupsd's problem. edit: bug report created by gnomuz in 2020 here:

New cpu installed. i'll try to run a parity check now and see what happens.

I started the RMA process and should receive the replacement CPU within the week.. I tried disabling docker and VMs and just running mover for the past 24 hours and it did not result in a crash, i turn on VMs (HAOS) and it results in a "hung" state within 4 hours. This time i was able to grab diagnostics since i destroyed the zpool. Not sure if it shows anything or if it's a different issue altogether.

was a long shot and it didnt work. I cant really think of anything else to try... It does seem like the server crashes more frequently now so that might hint that the CPU is even further degraded? the only error message i'm receiving is fro the IPMI but that doesn't really tell me what the issue is... Any other ideas @JorgeB?

I have an inkling that it's related to my HA VM. i think i've had it twice now that i trigger a lightswitch and shortly thereafter the server crashes. it might be the "USB manager" plugin i use to add a zwave and zigbee dongle to the same VM even though they have the same name in unraid, ill just add them in xml instead and see.

different PSU... still hard crash after 30~ hours 😬

Another full crash after 2 days 3 hours of uptime, no special activities were taking place (no mover, no manual moves, no parity check). I did get a troubling message yesterday about the 3.3V voltage sensor going off after searching the logs it appears i got a warning on the 28th of july as well, this time on the 12V voltage sensor I do have an extra PSU lying around, ill change it out and see if that fixes it...

i just ran into the same issue. setting the slimSAS port to SATA and plugging in a cable, the nvme slots AND all the pci slots stop working (even the ipmi board in the x1 slot stops working) Did you ever get it figured out with slimsas -> sas backplane cable? i was hoping to avoid getting a second HBA card if i can. i wasn't aware that these cables have any directionality to them and i am unable to find a "reverse breakout" SFF-8654 to SFF-8087 cable. this is the cable i got: https://www.inline-info.com/en/products/cable/sas/11195/inline-slim-sas-kabel-sff-8654-zu-mini-sas-sff-8087-12gb/s-0.5m edit: the manufacturer responded to my inquiry so it would appear they DO have directionality, i'll see if i can find such a cable on ebay instead...

The server ran fine for 4 days and 12 hours~ and then out of nowhere it did another hard crash where keyboard, webui, ssh, containers unresponsive, etc. Had to do a hard reset Syslog for the past couple hours: The crash happened a couple of minutes after i finished manually moving a couple of things (400GB~) off of the cache to the array. The mover is disabled. the self test didn't find any errors either. Could it be a bad PSU? edit: just saw my motherboard got ANOTHER BIOS update (beta 3802) though it specifically mentions a fix for non-k processors - I'll install that as well and see what happens...

I remember from the old docker folders plugin there was an option to indicate if a container with autostart enabled is not running, the green play symbol would turn into a yellow pause symbol. This to indicate that a container with autostart enabled is no longer running. This was very useful to me since i have some big folders and its hard to see at a glance if something crashed or not. Is it possible to add this function to folderview?

Since it seemed like the mover was causing the crash i searched around for previous issues pertaining to the mover crashing the server. I tried running xfs repair on all the disks in maintenance mode, a couple of them seem to have needed that with 10~ inodes being fixed. running "docker safe new perms" which seemed to get stuck at /disk13/Storage for over 15 hours before i decided cancel it. increasing the setting "minimum free space" for all my shares, i believe i set them when i created the server and didn't touch since. I then enabled mover logging and it ran for 5 minutes before crashing the server, stopping on disk13/appdata/Radarr/MediaCover/2018/poster-500.jpg so disk13 is mentioned in 2 issues, might be a coincidence. the disc is only about 3 months old and doesn't seem to have any obvious issues 1 Raw read error rate 0x000b 100 100 001 Pre-fail Always Never 0 2 Throughput performance 0x0004 148 148 000 Old age Offline Never 49 3 Spin up time 0x0007 094 094 001 Pre-fail Always Never 145 (average 137) 4 Start stop count 0x0012 100 100 000 Old age Always Never 62 5 Reallocated sector count 0x0033 100 100 001 Pre-fail Always Never 0 7 Seek error rate 0x000a 100 100 000 Old age Always Never 0 8 Seek time performance 0x0004 140 140 000 Old age Offline Never 15 9 Power on hours 0x0012 100 100 000 Old age Always Never 2610 (3m, 17d, 18h) 10 Spin retry count 0x0012 100 100 000 Old age Always Never 0 12 Power cycle count 0x0032 100 100 000 Old age Always Never 62 22 Helium level 0x0023 100 100 025 Pre-fail Always Never 6553700 90 NAND master 0x0031 100 100 001 Pre-fail Offline Never 0x004400000000 192 Power-off retract count 0x0032 100 100 000 Old age Always Never 214 193 Load cycle count 0x0012 100 100 000 Old age Always Never 214 194 Temperature celsius 0x0002 049 049 000 Old age Always Never 45 (min/max 25/56) 196 Reallocated event count 0x0032 100 100 000 Old age Always Never 0 197 Current pending sector 0x0022 100 100 000 Old age Always Never 0 198 Offline uncorrectable 0x0008 100 100 000 Old age Offline Never 0 199 UDMA CRC error count 0x000a 100 100 000 Old age Always Never 0 I'm currently running an extended self-test but im doubtful that will tell me anything. Anything else i can do to troubleshoot Mover if that indeed is the issue?

i updated to 6.12.11 when it was released and my server started crashing this friday so i asssume that is not the issue, i will most likely do a rollback if my other troubleshooting does not fix the problem.

Formatted the cache as btrfs and started mover to try to move stuff back over... instantly crashed would appear that the issue is with either unraid or my storage...