Hey Everyone!

I am having a couple issues and hoping to get some help as I seem to have hit a wall.

The first couple issues I am having may be related, but I would say the first one is my 'high priority' issue:

1. My system boots and I have it set to start my array automatically at startup. If I then go and stop the array, the array wont start again. Syslog shows 'already starting' or something similar if I click the 'start' button more than once. I have given it as long as 20 or 25 minutes to start before, no luck. Rebooting will then start the array, which is the 'workaround'

2. Possibly related to this, but I cant say for sure. When I do reboot, it is taking approximately 5 minutes to reboot which seems like a really long time.

Any suggestions would be appreciated. I've attached some of the diagnostics data.

unraid-diagnostics.zip

Those diagnostics are incomplete. Did you edit them?

What can you tell us about your remote mount? Is it set to automount?

4 minutes ago, trurl said:

What can you tell us about your remote mount? Is it set to automount?

2 remote mounts. SMB. Neither set to automount.

You didn't mention disk5 was disabled, and not immediately obvious since you removed the smart folder from those diagnostics.

Just now, trurl said:

You didn't mention disk5 was disabled, and not immediately obvious since you removed the smart folder from those diagnostics.

Thats one of the other issues that I'm having, that I didnt mention here because I honestly didnt think it was related. But the issue with disk5 is this - I had it in 1 chassis. A jonsbo.... something?? But it was getting hot in there because the jonsbo cases pack the drives right next to each other. So I took disk5 out of there and was moving it to a shelf networked via SAS. but I need to use an interposer and therefore unraid was not recognizing it as the same disk. so ive gone through all sorts of headache with disk5 too - and I was potentially going to post something about it later. but i didnt think it was related to this so was hoping to get this resolved before addressing that.

but if we want to hop into it all at the same time, yes disk5 is disabled because of the above.

Is disk5 attached currently? Can't really tell since your diagnostics are incomplete.

42 minutes ago, tjm said:

When I do reboot, it is taking approximately 5 minutes to reboot which seems like a really long time.

Not necessarily unusual.

43 minutes ago, tjm said:

If I then go and stop the array, the array wont start again. Syslog shows 'already starting' or something similar if I click the 'start' button more than once. I have given it as long as 20 or 25 minutes to start before, no luck. Rebooting will then start the array, which is the 'workaround'

Do you have a syslog that shows that?

Setup Syslog Server

4 minutes ago, trurl said:

Can't really tell since your diagnostics are incomplete.

Diagnostics are anonymized if you tell it to. Only those user shares defined by docker.cfg and domains.cfg are named completely. And your disk serial numbers are in the logs and in system/vars.txt anyway so already in the diagnostics you posted. Don't see any reason to leave out what you did.

alright, i apologize. i thought i was being more thorough. please see attached for complete diagnostics.

but to answer your question, no disk5 is not currently attached.

when i was trying to attach it, i was trying to run xfs_repair with a few different flags but it kept coming back with error 117

so i plugged it into a windows machine and it didnt see anything on the drive, so im running UFS Explorer, which did find a bunch of the now-missing data and am copying it back onto the array

unraid-diagnostics2.zip

3 minutes ago, trurl said:

Do you have a syslog that shows that?

i can get you one, but ufs explorer is copying right now so it may be a couple hours before i can send you this

Just now, tjm said:

when i was trying to attach it, i was trying to run xfs_repair with a few different flags but it kept coming back with error 117

so i plugged it into a windows machine and it didnt see anything on the drive, so im running UFS Explorer, which did find a bunch of the now-missing data and am copying it back onto the array

Disk 5 is being emulated by parity just fine, it is mounted, and contains 6.3T of data. You can browse its files without even bothering with the physical disk. And if you rebuild it, the emulated contents are exactly what you will get.

Did you check filesystem on the emulated disk5, or what did you do?

Why do you think anything is missing?

Just now, trurl said:

Disk 5 is being emulated by parity just fine, it is mounted, and contains 6.3T of data. You can browse its files without even bothering with the physical disk. And if you rebuild it, the emulated contents are exactly what you will get.

Did you check filesystem on the emulated disk5, or what did you do?

Why do you think anything is missing?

when i went to move disk5 from the first machine to the shelf, i had to use interposers as mentioned.

so at first, unraid started doing Disk-Rebuild which I believe is correct and as expected.

but, sometime between 50% and 80% through the Disk-Rebuild process, Unraid silently crashed where I was trying to go to the url and it was telling me 'no connection'.

in fact, its been silently crashing a lot. and it looks to be Docker related since I disabled Docker entirely about 5 days ago and it hasn't happened since.

but yeah, silently crashed and then I was asking AI what to do and she pointed me in the direction of xfs_repair but even when forcing it, it was throwing that 117 error. so shes the reason why im ultimately on UFS Explorer and copying things.

if there was a better way for me to have done this, please I am all for learning. thanks!

Regarding the silent crash, if it helps.

* I did mirror syslog onto flash for a while there, but there would be hours without any events.

For example, if it were to crash right now (1545 14FEB2026), syslog would (of course) show information from me restarting it right now, but the last event prior to that was around 0800 13FEB2026.

And I cant plug her in to a monitor right now.

Still not clear whether you were trying to run xfs_repair on the emulated disk, or on the physical disk. And were you trying to use the command line, or the webUI to run it. Easy to get the command line wrong in more than one way.

Unless emulated disk5 was showing unmountable, I would say AI advice was a complete failure.

17 minutes ago, tjm said:

if there was a better way for me to have done this

23 minutes ago, trurl said:

Disk 5 is being emulated by parity just fine, it is mounted, and contains 6.3T of data. You can browse its files without even bothering with the physical disk. And if you rebuild it, the emulated contents are exactly what you will get.

And these latest diagnostics don't show any lost+found share, so I doubt you are missing anything.

Getting disk5 rebuilt is the way forward for your data.

Have you done memtest recently?

5 minutes ago, trurl said:

Still not clear whether you were trying to run xfs_repair on the emulated disk, or on the physical disk. And were you trying to use the command line, or the webUI to run it. Easy to get the command line wrong in more than one way.

Unless emulated disk5 was showing unmountable, I would say AI advice was a complete failure.

Command Line, but it is very likely that AI was giving me the wrong advice

1 minute ago, trurl said:

And these latest diagnostics don't show any lost+found share, so I doubt you are missing anything.

Getting disk5 rebuilt is the way forward for your data.

Have you done memtest recently?

Alright, so it sounds to me like I should stick disk5 back in and Disk-Rebuild should run?

No, I have not done memtest recently as this is only doable during boot, right?

Since I cant hook it up to a monitor right now, I have no way to catch it during boot.

Maybe not as thorough as bootable memtest

41 minutes ago, tjm said:

ufs explorer is copying right now

Copying to where?

Just now, trurl said:

Copying to where?

to array

44 minutes ago, tjm said:

plugged it into a windows machine and it didnt see anything on the drive, so im running UFS Explorer

Windows can't natively read any Linux filesystem. UFS Explorer certainly can, but in this case probably any Linux could have read it including Unraid.

14 minutes ago, tjm said:

to array

So if you rebuild disk5 you will have duplicates?

55 minutes ago, tjm said:

* I did mirror syslog onto flash for a while there, but there would be hours without any events.

And I cant plug her in to a monitor right now.

Setup syslog server to log to a user share that you can access over the network, then you can zip it up and post it.

It is also possible to access flash drive over the network if you share it.

12 minutes ago, trurl said:

So if you rebuild disk5 you will have duplicates?

thats a good question. i imagine the answer is yes. but i can deal with that later.

Alright, following back up on this.

Disk5

I have attempted to add this back to the array. It says "All existing data on this device will be OVERWRITTEN when array is Started" which I think is fine (?). So I am attempting to start the array

Array not Starting

To add disk5 back in, I stopped the array and now I am attempting to start it. It won't start. Clicking the 'Start' button more than once causes the page to refresh and it 'jumps' to the top of the 'Main' tab, but nothing else happens.

This is what is shown in syslog:

But I've given it as long as ~20 minute before and it does not start.

Live MemTest

I ran this on about half the installed ram, and it came back without any errors.