Hi,

I'm greeted on the Docker Tab as the title states: "Docker Service failed to start".

I saw that I accidentally filled the whole cache pool and thought.... "Well, I guess that's the problem", cleared up some space and... nope, still failed to start. Then I checked the Log and saw that there were some errors with sdb (which makes me sweat a little, since it's one of my 14TB drives) and BTRFS errors on sdc, which would make more sense in the context of Docker, since sdc should be my cache (pool) and apps/docker are on there. sdc and sdf are some "old" 250GB SSDs which I use for the pool. First I thought it was caused by sdf, since the SATA cable came loose a bit(? -> unRAID reporting the SSD as disconnected) when I opened my case to install an Intel Gigabit CT Desktop NIC, but I guess that was just coincidence? The array wasn't on while I was doing this.

How should I go about this? I don't dare to try anything in repairing BTRFS, without someone's more experience opinion...

tower-diagnostics-20260202-2156.zip

Looks like problems with connections to cache and disk3. After those are fixed we can see what else might need to be done.

24 minutes ago, trurl said:

Looks like problems with connections to cache and disk3. After those are fixed we can see what else might need to be done.

Well, I thought I did that... checked all cables and no disk is shown as error / no CRC errors pop up.

Lots of this in syslog

Feb  2 21:40:28 Tower kernel: ata1.00: status: { DRDY }
Feb  2 21:40:28 Tower kernel: ata1: hard resetting link

ata1

[1:0:0:0]    disk    ATA      SAMSUNG MZ7LN256 1H6Q  /dev/sdc   /dev/sg2 

And lots of this

Feb  2 21:40:17 Tower kernel: ata5.00: status: { DRDY }
Feb  2 21:40:17 Tower kernel: ata5: hard resetting link

ata5

[5:0:0:0]    disk    ATA      WDC WD80EMZZ-11B 0A81  /dev/sdg   /dev/sg6 

Sorry, should be cache2 and disk6

ata1

[0:0:0:0]    disk    ATA      CT240BX200SSD1   .6    /dev/sdf   /dev/sg5 
  state=running queue_depth=32 scsi_level=6 type=0 device_blocked=0 timeout=30
  dir: /sys/bus/scsi/devices/0:0:0:0  [/sys/devices/pci0000:00/0000:00:1f.2/ata1/host0/target0:0:0/0:0:0:0]

ata5

[4:0:0:0]    disk    ATA      TOSHIBA MG07ACA1 0103  /dev/sdb   /dev/sg1 
  state=running queue_depth=32 scsi_level=6 type=0 device_blocked=0 timeout=30
  dir: /sys/bus/scsi/devices/4:0:0:0  [/sys/devices/pci0000:00/0000:00:1f.2/ata5/host4/target4:0:0/4:0:0:0]

Edit: uh... why the hell is all my text gone?! Have to retype this when I'm back.

I will keep the diag attached. Reattached the cables and didn't see the errors anymore.

tower-diagnostics-20260203-1757.zip

Edited February 3Feb 3 by GerryGER

15 minutes ago, GerryGER said:

Reattached the cables and didn't see the errors anymore.

That looks OK now.

Feb  3 17:53:04 Tower kernel: BTRFS error (device sdc1): level verify failed on logical 1528780898304 mirror 2 wanted 1 found 0

Run scrub on cache pool.

All of your disks are too full. Some work to be done there and with your user shares.

You will have to recreate docker.img, we will get to that later. In the meantime, disable Docker and VM Manager in Settings and leave them disabled until we work on the other stuff.

2 hours ago, trurl said:

All of your disks are too full. Some work to be done there and with your user shares.

Yeah, with current HDD prices, that is only going to change marginally. Maybe I will gut my 4TB external and use that for the time being.

2 hours ago, trurl said:

You will have to recreate docker.img, we will get to that later. In the meantime, disable Docker and VM Manager in Settings and leave them disabled until we work on the other stuff.

Disabled both, I'm not using VMs anyway.

Scrub:

2 hours ago, trurl said:

All of your disks are too full.

I said "too" full, because you are asking for trouble.

3 minutes ago, GerryGER said:

that is only going to change marginally

Find some stuff to delete.

Reboot and post new diagnostics.

6 minutes ago, trurl said:

I said "too" full, because you are asking for trouble.

Probably even the cause of the trouble that made you start this thread, and other trouble is possible.

1 hour ago, trurl said:

Find some stuff to delete.

Yeah, will do. ;)

1 hour ago, trurl said:

Reboot and post new diagnostics.

Attached.

tower-diagnostics-20260203-2227.zip

You do at least have Minimum Free set for each of your user shares, though smaller than usually seen (or recommended).

I don't recommend sharing disks on the network.

You have a lot of .cfg files in config/shares for user shares that no longer exist.

Go to the User Shares page and click CLEAN UP button at the bottom.

Click on cache to get to its page and tell me what it has for Minimum free space.

23 hours ago, trurl said:

I don't recommend sharing disks on the network.

Must have been an old .cfg as they are definitely NOT shared. Not in unRAID and not visible on the network.

23 hours ago, trurl said:

Go to the User Shares page and click CLEAN UP button at the bottom.

Never saw that option, thanks, just did that.

23 hours ago, trurl said:

Click on cache to get to its page and tell me what it has for Minimum free space.

I now put it at 15GB min. and cleared up 50GB on the Cache. A few 100GB on the Array itself. As soon as Docker runs again, I can clean up more stuff.

So, can I consider everything ok again? What's the safest way to restoring Docker? Deleting it via the WebGUI and restoring the containers? (They should be recoverable via the Docker selection or not?)

https://docs.unraid.net/unraid-os/troubleshooting/common-issues/docker-troubleshooting/