Can't start any Docker Container in Unraid

[Blizii]

I have a Unraid Server and i have noticed that i can't start a pterodactyl server anymore. So i restarted my Server. Now i can't start any Docker Container even in the Unraid UI. Please Help me!

 

Examples:

Cloudflared: docker: Error response from daemon: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: exec: "cloudflared": executable file not found in $PATH: unknown.

 

Minecraft Server: docker: Error response from daemon: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: exec: "/start": permission denied: unknown.

 

Minecraft Server But i can't remember what i have done with permissions. I also tried this command on my Gameserver Share: "chmod -R 777 *"

[trurl]

attach diagnostics to your NEXT post in THIS thread.

[Blizii]

This are the Diagnostics:

server01-diagnostics-20230112-1642.zip

[trurl]

Your appdata, domains, and system shares are all set to prefer a pool named 'cache', which you don't seem to have now, so these have been recreated on the array.

 

Looks like you had pools named 'cache_appdata' and 'cache_storage', but these don't exist now.

 

The only pool you have currently is 'gameservers',  which does have part of appdata on it, probably because you mapped it directly instead of the appdata user share.

 

And no attached disks that might have been used for other pools.

 

In fact, FCP is giving you warnings about these very things

Jan 12 16:34:21 Server01 root: Fix Common Problems: Warning: Share appdata references non existent pool cache
Jan 12 16:34:21 Server01 root: Fix Common Problems: Warning: Share domains references non existent pool cache
Jan 12 16:34:21 Server01 root: Fix Common Problems: Warning: Share isos references non existent pool cache
Jan 12 16:34:21 Server01 root: Fix Common Problems: Warning: Share system references non existent pool cache

You should fix those and the many other FCP warnings you are getting. Except, leave those docker updates for later until you get the other things fixed since docker isn't working and you should not even try to use it until things are cleaned up.

 

Jan 12 16:25:48 Server01 kernel: md: recovery thread: recon D3 ...

Go to Settings and disable Docker and VM Manager and finish rebuilding disk3. Leave them disabled until you have cleaned up this mess.

 

What do you want to do with all this since you only have one pool now? You want appdata, domains, system shares to all get moved to that 'gameservers' pool? Or what?

 

 

[Blizii]

No, currently i don't use any Pool for all Shares except the Gameservers Share, because i don't can find a HBA Card that works with Unraid, so i don't have enough Ports for Cache Devices.

 

Where can i see all the FCP Warnings?

 

And on the attached Picture you can see, i don't have assigned a pool to the other shares, why is that error there?

 

Sorry I'm new to unraid and still don't quite understand a lot of things.

[trurl]

11 minutes ago, Blizii said:

Where can i see all the FCP Warnings?

I saw them in your syslog. You can see them in Settings - Fix Common Problems. Have you not been getting Notifications about them?

 

18 minutes ago, trurl said:

Go to Settings and disable Docker and VM Manager and finish rebuilding disk3. Leave them disabled until you have cleaned up this mess.

Have you done that yet?

 

Does that screenshot have all of your user shares? I don't think so since it is only showing user shares with Capitalized names. I don't recommend using Capitalized names for anything. You must be very careful about upper/lowercase. Linux is. Appdata and appdata are not at all the same thing for example.

 

What do you get from the command line with this?

ls -lah /mnt/user

 

[Blizii]

1 minute ago, trurl said:

Have you done that yet?

Yes, it's not finished but already disabled. And the Names are only in the UI Capitalized. I attach the picture with the Output.

 

4 minutes ago, trurl said:

Have you not been getting Notifications about them?

No i only can remeber about Notifications for Docker Updates.

 

 

 

[trurl]

Go to User Shares, click on appdata, and post a screenshot if its Share Settings.

[Blizii]

I fixed the Problem with the pools on the system, domains, isos and appdata Shares. I have to set it to "prefer", apply and change it back to "off" that it really applies.

 

[Blizii]

This are the new diagnostics after fix the share problems, and updated all plugins.

The rebuild is at 17%

 

server01-diagnostics-20230112-1742.zip

[trurl]

You should set domains and system shares to prefer also. Then after rebuild completes we can run mover to get these shares moved to their designated pool.

 

plugins

 

ca.backup2.plg - 2022.12.13  (Deprecated)  (Up to date)

There is a new version of this but you have to uninstall and then install the new one from Apps. You might want to look at the thread for the new one:

https://forums.unraid.net/topic/132721-plugin-ca-appdata-backup-restore-v3/

 

NerdPack.plg - 2021.08.11  (Up to date)  (Incompatible)

You installed the replacement NerdTools, but you didn't remove the incompatible NerdPack

 

tasmotapm.plg - 2021.03.24  (Unknown to Community Applications)

I don't know what this is (was?). Do you need it?

 

[Blizii]

Okay, i now removed the three Plugins. The third was a test for my Tasmota smart plug.

 

Now has the appdata, domains and system Share the Cache to prefer.

 

And yes that was another question, i don't really understand how mover works.

 

Do I have to wait for the rebuild now?

[trurl]

21 minutes ago, Blizii said:

Do I have to wait for the rebuild now?

Not technically but disk access will affect rebuild performance and vice versa. And if rebuild fails things could get more complicated.  Since there will probably be a lot to move I would wait 

[Blizii]

Okay, I think this will take a few more hours, I'll write to you as soon as it's ready so we can continue with that. Nevertheless, thank you very much.

[Blizii]

Okay the data rebuild is now finished. What have i to do now?

[trurl]

Run mover, wait for it to complete, post new diagnostics.

 

I will be out for a few hours, will check back later.

[Blizii]

see attachments

 

server01-diagnostics-20230113-2057.zip

[trurl]

appdata, domains, system still have files on the array because your gameservers pool ran out of space.

 

What is the purpose of this prefer share

g--------r                        shareUseCache="prefer"  # Share exists on gameservers

 

 

Also, you are having connection problems on disk2. See if you can fix those before we work on the other things.

[trurl]

Go to User Shares page, click Compute All button at bottom, wait for the complete results, post a screenshot

[Blizii]

- The g---r Share is for my Gameservers because they need a SSD to work.

- I didn't can find any problems at the Hardware on disk 2. (I have a Server Case with Hotswap)

And i have clicked on compute all. The Screenshots are down below.

 

 

 

Edited January 14 by Blizii
added screenshots

[trurl]

15 hours ago, trurl said:

post a screenshot

 

[Blizii]

I have the screenshot added in the message above.

[trurl]

3 hours ago, Blizii said:

I have the screenshot added in the message above.

Just a note about using the forum. If you edit a previous post that has already been read, nobody will know there is anything new to see. Your thread might have been forgotten if you hadn't made this new post.

 

 

[trurl]

Looks like most of appdata did get moved to your pool. domains still has a little something on disk3, might just be an empty folder. system share still has a lot on disk1, maybe a large docker.img you aren't using anymore?

 

What do you get from command line with this?

ls -lah /mnt/disk3/domains

and this?

ls -lah /mnt/disk1/system

 

4 hours ago, Blizii said:

- The g---r Share is for my Gameservers because they need a SSD to work.

If you really need all that your pool isn't large enough for the other shares. You might consider adding a separate pool and put appdata, domains, system there. Maybe you already had something like that before since you still have .cfg for other pools.

 

 

[Blizii]

20 hours ago, trurl said:

ls -lah /mnt/disk3/domains

 

20 hours ago, trurl said:

ls -lah /mnt/disk1/system

 

20 hours ago, trurl said:

If you really need all that your pool isn't large enough for the other shares. You might consider adding a separate pool and put appdata, domains, system there. Maybe you already had something like that before since you still have .cfg for other pools.

My Goal is to have for the System, Appdata and Domains a seperate pool, but my problem is, i don't find a good hba card. And until i have a hba card i don't have enough ports on my motherboard.

20 hours ago, trurl said:

Just a note about using the forum. If you edit a previous post that has already been read, nobody will know there is anything new to see. Your thread might have been forgotten if you hadn't made this new post.

Okay thank you, i will not doing this anymore.