I've had my unraid server running for a very long time and never make any changes to it other than keeping up to date with the latest version. Currently running 7.3.2

For the past year or so it restarts or crashes and ends up running a parity check everytime. For the life of me I can't figure out why? I have nothing sceduled to tell it to do this or any scripts running at all. I only have 4 data drives and a parity drive, that all have been replaced with wd red drives in the last 6 months. The server was doing the same thing even before I updated the drives. I have checked the server logs and there is no errors showing as to why it would restart at the same time every week? I am only running 1 docker, and it is the plex docker. I only use my server for watching video files locally and data backup. I have updated some hardware in my server, eg. ram, and psu. I have ran memcheck on my memory and have not discovered any errors? Not sure what other info to share here for a possible solution to this issue?

tower-diagnostics-20261005-2113.zip

Edited October 6Oct 6 by GrahamW

The diagnostics include the log from before the last event, which helps.

The server was running normally until about 03:40 on Monday, October 5. The last entries are the daily update check and Mover, both routine. After that there is nothing, no crash, hardware error, disk error or shutdown. The next boot is at 10:51, and Unraid detected an unclean shutdown, which is why the parity check started. In short, the server stopped suddenly without logging anything.

When this happens on the same day every week and nothing in Unraid is scheduled for that day, the first thing to check is the UPS. Many UPS models run an automatic battery self-test, often once a week. During the test the UPS moves the load to battery. If the battery is old or weak, the output can drop for a moment and the server loses power. That leaves nothing in the log. It would also explain why new RAM, PSU and disks made no difference.

I see an Eaton Ellipse PRO connected. Can you check the following?

1. How old is the UPS battery? Can you see the automatic self-test schedule or event log on the UPS or in Eaton's software? Do the test times match the day of the week this happens?

2. As a test, connect the server directly to the wall, or to a different UPS, for the next two or three weeks. If the problem stops on that day, the UPS is the cause.

3. When you find the server after an event, is it switched off, on but not responding, or already restarted? Which day of the week is it usually, and is it usually in the early morning?

4. Your syslog is also sent to a remote syslog server. Check that log for October 5 between 03:40 and 10:51 and post anything you find there.

Also, the built-in UPS service (apcupsd) is running. It is made for APC units and may not talk correctly to an Eaton UPS. NUT is installed but not running. If Settings > UPS Settings does not show a correct battery charge and status, disable the built-in UPS service and set up NUT for the Eaton. Then a power event will be logged, and the server can shut down cleanly when the battery runs low.

Until this is solved, keep the remote syslog enabled, or turn the syslog mirror to flash back on, so the next event leaves a log.

On 10/5/2026 at 11:47 PM, JorgeB said:

The diagnostics include the log from before the last event, which helps.

The server was running normally until about 03:40 on Monday, October 5. The last entries are the daily update check and Mover, both routine. After that there is nothing, no crash, hardware error, disk error or shutdown. The next boot is at 10:51, and Unraid detected an unclean shutdown, which is why the parity check started. In short, the server stopped suddenly without logging anything.

When this happens on the same day every week and nothing in Unraid is scheduled for that day, the first thing to check is the UPS. Many UPS models run an automatic battery self-test, often once a week. During the test the UPS moves the load to battery. If the battery is old or weak, the output can drop for a moment and the server loses power. That leaves nothing in the log. It would also explain why new RAM, PSU and disks made no difference.

I see an Eaton Ellipse PRO connected. Can you check the following?

1. How old is the UPS battery? Can you see the automatic self-test schedule or event log on the UPS or in Eaton's software? Do the test times match the day of the week this happens?

2. As a test, connect the server directly to the wall, or to a different UPS, for the next two or three weeks. If the problem stops on that day, the UPS is the cause.

3. When you find the server after an event, is it switched off, on but not responding, or already restarted? Which day of the week is it usually, and is it usually in the early morning?

4. Your syslog is also sent to a remote syslog server. Check that log for October 5 between 03:40 and 10:51 and post anything you find there.

Also, the built-in UPS service (apcupsd) is running. It is made for APC units and may not talk correctly to an Eaton UPS. NUT is installed but not running. If Settings > UPS Settings does not show a correct battery charge and status, disable the built-in UPS service and set up NUT for the Eaton. Then a power event will be logged, and the server can shut down cleanly when the battery runs low.

Until this is solved, keep the remote syslog enabled, or turn the syslog mirror to flash back on, so the next event leaves a log.

Thank you for this valuable information. I never even considered the ups as being a problem. I think the battery inside it is in need of replacement. I have removed the ups for now to see if this solves the issue. Thanks again for this