Execution Error on starting/stopping dockers.

[trurl]

1 minute ago, ridley said:

8-15 you mean? 

I removed them and replaced them with larger drives after a bit of unbalancing.

Your diagnostics show disk15 mounted.

 

I don't understand what you mean though. If you removed the disks then you didn't replace them. Could you explain that better?

[ridley]

17 minutes ago, trurl said:

Still a lot of appdata all over your array, and system also still on array disks.

 

Even though it is mounted, corruption logged on disk24.

 

Many of your disks are very full. Why don't you have dual parity with so many disks?

 

Post a screenshot of Array Devices

 

 

 

 

I am trying to reduce my number of disks down to 8 in total so I do not need to power up the diskshelf. So the 3 , 4Tb  and some of the 8Tb drives will be going. I have a 14Tb and another 18Tb to add.

What sort of corruption on disk24?

 

I have looked at what is on the other appdata folders on non cache drives and they appear to be (almost) empty, just a folder structure. 

The system folder on Disk 2 is a folder called docker with 400Gb docker disk image file in it, the docker folder on the cache drive is empty.


 

Edited August 15, 2022 by ridley

[trurl]

15 minutes ago, ridley said:

system folder on Disk 2 is a folder called docker with 400Gb docker disk image file

Go to Docker in Settings, delete docker.img. Leave docker disabled for now. You will be recreating docker.img later as 20G, and it will be put on cache where it belongs.

18 minutes ago, ridley said:

reduce my number of disks down to 8

So what you meant was you have been replacing some drives with larger drives, and shrinking the array after moving files off other drives then removing them. Is that it? What method did you use to remove the drives? New Config and rebuild parity, or New Config after clearing the disks?

 

Post a screenshot of Compute... for appdata.

 

Then check filesystem on disk24

 

 

[ridley]

49 minutes ago, trurl said:

 

So what you meant was you have been replacing some drives with larger drives, and shrinking the array after moving files off other drives then removing them. Is that it? What method did you use to remove the drives? New Config and rebuild parity, or New Config after clearing the disks?

 

 

 

 

 

I cannot remember now, it was some time ago. I think it was the former.

[ridley]

2 hours ago, trurl said:

Go to Docker in Settings, delete docker.img. Leave docker disabled for now. You will be recreating docker.img later as 20G, and it will be put on cache where it belongs.

So what you meant was you have been replacing some drives with larger drives, and shrinking the array after moving files off other drives then removing them. Is that it? What method did you use to remove the drives? New Config and rebuild parity, or New Config after clearing the disks?

 

Post a screenshot of Compute... for appdata.

 

Then check filesystem on disk24

 

 

 

Appdata Compute, checked file system on 24 but it has errors

 

 

Edited August 15, 2022 by ridley

[ridley]

Disk 24 File Check

 

 

[trurl]

6 hours ago, ridley said:

appdata folders on non cache drives

What do you get from command line with this?

ls -lah /mnt/user0/appdata

 

[ridley]

14 hours ago, trurl said:

What do you get from command line with this?

ls -lah /mnt/user0/appdata

 

 

There you go.

I have repaired the file structure on disk25 and re run the mover and this is what I get,

 

what data is on the non cache drives appdata is a mixture of 

binhex preclear
krusader
Dolphin
binhex plexpass
kde
binhex krusader

 

The krusader ones can just be reinstalled, pre clear not needed. Plexpass if push came to shuff I could get the DB to regenerate it self., prefer not to if possible.

 

 

Edited August 16, 2022 by ridley
Changes

[trurl]

And what do you get from command line with this?

ls -lah /mnt/cache/appdata

 

[ridley]

11 minutes ago, trurl said:

And what do you get from command line with this?

ls -lah /mnt/cache/appdata

 

 

There you go 

 

[trurl]

As you can see, the appdata folders on the array are also appdata folders on cache.

  

On 8/14/2022 at 2:09 PM, trurl said:

mover won't move duplicates, so there may be some manual cleanup left to do.

 

[ridley]

7 minutes ago, trurl said:

As you can see, the appdata folders on the array are also appdata folders on cache.

  

 

 

So anything that's on the disks that is duplicated on the cache drive can be deleted? 

Can I delete completely empty folder structure on the disks without checking that they are on the cache drive? For instance on disk25 I have appdata/binhex-plexpass that has 16000 folders which contain 0 bytes...

Any tools that I can use to compare the cache drive to disks for the cleanup?

[trurl]

Dynamix File Manager plugin will let you examine each disk, and let you delete anything you don't need or move anything you need to keep. It pretty much eliminates the need for Krusader.

[ridley]

 

OK , what now?

 

Presumable rebuild the docker image?

 

 

[trurl]

post new diagnostics

[ridley]

there you go

 

tower-diagnostics-20220816-1910.zip

[trurl]

OK

36 minutes ago, ridley said:

rebuild the docker image

change size to 20G

 

https://wiki.unraid.net/Manual/Docker_Management#Re-Create_the_Docker_image_file

 

https://wiki.unraid.net/Manual/Docker_Management#Re-Installing_Docker_Applications

[ridley]

BTRFS or XFS?

 

Why 20G when my old one is 400Gb in size?

Edited August 16, 2022 by ridley

[Squid]

1 hour ago, ridley said:

Why 20G when my old one is 400Gb in size?

Mainly because there's no real reason for it to be that size.  20G is more than sufficient for 99% of the users.  The other 1% 40G would suffice.

 

400G means that if everything is configured correctly you wind up wasting ~380G of space on the cache drive.  If you've had problems with the image constantly filling up and just can't make heads or tails of why, then instead of an image, use a folder (Settings - Docker with the service stopped).  At least then it'll grow and shrink accordingly and only take up the space it actually needs.

[ridley]

34 minutes ago, Squid said:

Mainly because there's no real reason for it to be that size.  20G is more than sufficient for 99% of the users.  The other 1% 40G would suffice.

 

400G means that if everything is configured correctly you wind up wasting ~380G of space on the cache drive.  If you've had problems with the image constantly filling up and just can't make heads or tails of why, then instead of an image, use a folder (Settings - Docker with the service stopped).  At least then it'll grow and shrink accordingly and only take up the space it actually needs.

 

Thank you so much for all your help, finger crossed things seem to be working again.

Now I need to set up appdata / VM backup

Edited August 16, 2022 by ridley

[trurl]

If you do find you are filling docker.img, you should try to figure out why instead of just making it larger or making it a folder.

 

The usual reason for filling docker.img is an application writing to a path that isn't mapped.

 

Not only do you not want an application writing into docker.img and filling it, you can't even access the data it writes into docker.img.