JRubenol

Completed! I think I was just over hammering a system that is too old to handle that many things. Building out my new system but I'm fully operational now. Thanks again!

Rebuild started in safe maintenance mode. What I can already see is about a 35% improvement in speed rebuilding the drive. Hoping this time tomorrow, I'm back. And, I'm already building my next server. This one has served its purpose more than enough (multiple owners over 12 years).

Doesn't maintenance mode put the drives in read-only meaning I can't do the data rebuild?

Sorry I'm not entirely tracking this. You said setup a syslog server (which I have done and it's writing to boot USB which I'm fine with). Did you get all of the logs you needed? I can try in Maintenance mode, that's something I haven't tried. I just had another freeze (not rebuilding, that was paused, I was moving some stuff via ssh and nextcloud was syncing some stuff over samba).

I thought the post (two ago) that has syslog and syslog-previous is a syslog server....I have that enabled and writing to my boot drive (USB). Those are the logs. I'm not writing that much, a bit but yes, the rebuild hasn't finished. I currently have it paused and the server is still crashing. My syslog screen is attached.

Diagnostics attached. tower-diagnostics-20260407-1516.zip

And, another crash (not during rebuild, but, during moving files). Common theme seems to be SMB stuff mixed with other things. Lots attached. Diagnostics to follow. syslog syslog-previous

Good news is I got a very clean smartctrl test on it. Rebuilding now....if this fails, I'm starting to run out of ideas of what it could be. I have syslog server up and running so we'll see. hard_drive.txt

Going to run a long smart test first. It's running now. Going to take until 2am. Figured, might as well know where things stand completely. Will report back. Also, I'm still interested if you have a donation/buy me a coffee link.

Here's that diagnostics with disk4 removed from the array. Also - do you have a "buy me a coffee" link or anything? You're spending a lot of time with me and I appreciate it. tower-diagnostics-20260406-0911.zip

Working on this now, possibly side issue happening in that it's showing an unclean shutdown despite the fact that I did a clean shutdown through the GUI....separate post? I'm getting nervous I'm on the brink of losing data.

I just took out each drive, used compressed air to clean it all out (there was a wire that was pinched a bit....) and put the drives back in. I'm booting up right now. Question is, as soon as I start the array, the drive rebuild will start. You're asking bfor a diagnostics before that. So, just to clarify, you want one before I start the array?

Here it is. tower-diagnostics-20260406-0739.zip

I see that it looks like my parity drive is having some issues.....going to try to reseat it.....I only have single disk failure available now, and obviously with disk4 rebuilding, I could be in trouble. But, happy to take any suggestions as I limp through trying to recover from this.

Got those logs....crashed again. Hopefully this helps in some way. syslog-previous tower-diagnostics-20260405-2034.zip

Unfortunately, it didn't work. My system locked up again during rebuild, on forced power off and power on, it started over again.

Installed and ran. Didn't find any problems.

I can 100% guarantee a crash if I watch something from the unraid server as I rebuild the data disk. It'll crash within minutes. I'm hesitant to purposefully do it as I try to hobble through a full rebuild..... Also, the server is stored in an electronic closet not with a monitor, for me to get it onto a monitor I have to take it down and put it in my living room. I'd love to figure this out, just not sure I have the time immediately and right now the goal is really the rebuild if I can get it done.

Interesting. Multiple sources say otherwise. Including some first hand accounts on Reddit saying they were surprised (and happy) that it didn't.

Hi All, From what I've read, a clean reboot should not result in restarting data rebuild yet it did for me. I'm limping through diagnosing hardware issue while trying to get my raid parity back. To do this, my best bet is for me to pause every few percent of rebuild and do a clean reboot (if it were to not start over). Steps: Start data rebuild Let it go a few percent Pause data rebuild Clean reboot Result: on reboot, I had to manually start the array, rebuild started automatically, but, at 0%. Diagnostics to follow.

Diagnostics attached. Data rebuild is going on now. It's taxing the system but, this server has been running for quite some time for me and it does what I need it to do. I just need to get my parity back so my data is safe. This has always happened during rebuilds, it crashes pretty quickly if I try to access my media while rebuild is happening. I let it run 36 hours (estimate was 24-26 hours for rebuild) and it failed for me (despite the fact that I left it alone for a day and a half). tower-diagnostics-20260402-1935.zip

Site says, but when system locks up, I get locked out of ssh....so a hard reboot is required. WebGUI not available Access via SSH, telnet, or direct console to run the diagnostics command. The zip file saves to /boot/logs. Always capture diagnostics before rebooting to keep logs intact.

Hi All, In Googling how to throttle rebuild I found some old post about speed_limit_max but some indications that this has been depracated. When I try I get this: root@Tower:~# echo 5000 > /proc/sys/dev/raid/speed_limit_max -bash: /proc/sys/dev/raid/speed_limit_max: No such file or directory My system is older it's crashing when I try to rebuild data on a larger drive I installed. I'd like to throttle it way down, even if it means the rebuild takes 4-5 days. Thoughts?

More a curiosity than anything. My admin console for tailscale doesn't show my server ("Tower") in devices but I can see it with my Manjaro client running tailscale status and I can access my smb shares remotely without an issue. I could have sworn that when I set up tailscale I could see the Tower in my admin console.

Sometimes, a full reboot just does the trick. Problem resolved.