First time back to the forums in a long time, so forgive lack of etiquette if any. If there is any info needed that i haven't provided by the end let me know and I will add it. So everything was working fine but I happen to get a bigger HDD and I popped it in for my parity drive. I then consequently moved every drive down 1 slot to accommodate the new drive. parity 2 went to disk 1, and disk 1 went to disk 2, etc. (all the way down to disk 6 with that disk coming out of the array). cache drives were never touched. Now Ive done this many times before and retained all my apps, plugins, and data.  But this time around, after clicking "new config" and choosing to save my cache drive config.  After doing so, and setting the array the way I wanted, the next day after the parity check successfully ended, all of my docker apps are gone. the docker service itself works, it just has no apps to list and is waiting for me to install. I dont want to go through the headache of installing and re-setting up every app. I have the docker image, and appdata, it just doesn't seem to find it anymore? it is set to /mnt/user/system/docker.img any help would be greatly appreciated. 

Edited January 30, 20197 yr by visdomr
Time correction and spelling

I think we'd better clear up some terminology before we can come to a mutual understanding of what you did.

 

When I talk about "slots" I mean the actual disk assignments, which Unraid keeps track of using the serial numbers. Is that what you mean also? Because this doesn't make any sense if that is what you mean by slot:

1 hour ago, visdomr said:

get a bigger HDD and I popped it in for my parity drive. I then consequently moved every drive down 1 slot to accommodate the new drive. parity 2 went to disk 1, and disk 1 went to disk 2, etc. (all the way down to disk 6 with that disk coming out of the array)

followed by New Config instead of any rebuilding is going to remove data from your array. Disk6 data is removed, and new disk1 (former parity) has no data.

I took a look at his system. Here's the diagnostics zip. He re-assigned every drive with a new config when he installed the new parity drive. I pointed docker at his old image which is now located in what unraid previously saw as disk 1, but is now disk 2, and docker straight up refused to start. Didn't have time to help him further today (I know OP IRL)

tower-diagnostics-20190129-1923.zip

Here's the log from when I tried to start docker when it was pointed at the old image.
 

Jan 29 19:29:40 Tower emhttpd: shcmd (34312): /usr/local/sbin/mount_image '/mnt/disk2/system/docker/docker.img' /var/lib/docker 60
Jan 29 19:29:41 Tower kernel: BTRFS: device fsid 5932acc8-3d13-469c-ac36-2392ce4dfcc1 devid 1 transid 5090890 /dev/loop3
Jan 29 19:29:41 Tower kernel: BTRFS info (device loop3): disk space caching is enabled
Jan 29 19:29:41 Tower kernel: BTRFS info (device loop3): has skinny extents
Jan 29 19:29:41 Tower kernel: BTRFS error (device loop3): bad tree block start 0 21993553920
Jan 29 19:29:41 Tower kernel: BTRFS error (device loop3): bad tree block start 0 21993553920
Jan 29 19:29:41 Tower kernel: BTRFS warning (device loop3): failed to read root (objectid=4): -5
Jan 29 19:29:41 Tower kernel: BTRFS error (device loop3): open_ctree failed
Jan 29 19:29:41 Tower root: mount: /var/lib/docker: wrong fs type, bad option, bad superblock on /dev/loop3, missing codepage or helper program, or other error.
Jan 29 19:29:41 Tower root: mount error
Jan 29 19:29:41 Tower emhttpd: shcmd (34312): exit status: 1

 

Edited January 30, 20197 yr by Starlord

Those diagnostics don't really answer any of my questions and the new config isn't in the log so I guess it happened before reboot.

 

Do you understand what I was saying above? Because what he described doesn't make any sense. He seemed to think this was a normal way to upgrade a disk.

 

If you New Config by assigning a former parity disk to the disk1 slot, then that new disk1 will not have any data on it. And if you New Config and remove the former disk6, the data from that disk is no longer in the array.

 

Maybe he means something completely different. If not, I think it is pointless to worry about dockers when he has removed some of his data by mistake.

Disk 6 was initialized into the array but did not have user data on it. But you are correct. That's exactly how he did the disk swap.

Edited January 30, 20197 yr by Starlord
engrish

Well if disk6 didn't have any data on it I guess nothing is lost, but please make sure he understands that what he did is not normal and would have removed data from the array if there had been any data on that last disk. I assume he formatted the new disk1.

 

His docker image is way too large. Anybody that thinks they need more than 20G for the docker image is doing something wrong with their docker apps.

 

Also, the system share is set to not use cache. It is currently scattered all over the array. It should be cache-prefer and live on the cache drive.

 

Disable the docker service and delete the docker image. Set the system share to cache-prefer and run mover. Then go to the Shares page and Compute... the system share and see if it all got moved to cache. Once that is take care of we can proceed to the next thing.