I've had the same set of dockers for quiet a while now, along with a VM. Last night I developed some issues with my host controller and the system was down for some time, until this morning when I was able to get it back up. One of the discs was bad and I'm currently rebuilding it.

 

Any reason why the dockers will not start and the VM will not launch? I am getting the Libvirt Service failed to start error.

 

Thank you.

tower2-diagnostics-20181118-1629.zip

Both the libvirt.img and the docker.img files are trashed.

 

Docker.img is easy to fix -> Settings - Docker - Disable the Service, Delete the Image, enable the service, Apps - Previous Apps, Check off all your apps you want and install multi.

 

libvirt.img though you pretty much have to recreate from a backup as it has all of the VM xml's etc. (Although you should be able to redo the VM's and reuse the existing vdisks)

 

This is all probably because of this on the cache drive

Nov 18 12:18:37 Tower2 kernel: BTRFS info (device sdh1): bdev /dev/sdc1 errs: wr 1429608, rd 1310707, flush 8711, corrupt 0, gen 0

But, since it's a cache-pool, I'm going to bow out and let the experts advise.

Edited November 18, 20187 yr by Squid

Than you. Just reinstalled the dockers and back to normal. Now, how do I get the VM back? Is there a backup that I can use?

Not unless you made a backup of the libvirt.img (or used something like the appdata backup plugin).

 

From a "not-knowing person", the 1.4M write errors and 1.3M read errors don't seem like acceptable, but I'm not the one to help on that.

I have CA Bckup / Restore Appdata. However, when I click on it there does not seem to be any way to initiate a restore. I know I've used the appdate restore in the past. I am not sure why I do not see it now. I also have my old cache disk sitting around.

 

I just looked and I have a Win 10 vdisk1 sitting in a backup folder. Can I use that to restore my VM? 

VM is back! I ran the backup and it seems to have restored fine!

 

Thank you very much!

One of your cache devices likely dropped offline in the past, the Crucial SSD, replace cables, run a scrub a monitor for future issues.

12 hours ago, johnnie.black said:

One of your cache devices likely dropped offline in the past, the Crucial SSD, replace cables, run a scrub a monitor for future issues.

Thank you. I looked a little at the scrub but cannot find a guide how to use it. Any suggestions? The closest I came to was:

 

btrfs check --readonly --check-data-csum /dev/sdc1

 

 

It looks like you have two threads open, each dealing with a different set of problems.

 

 

Seriously, sort out your controller/disks/parity problems before worrying about dockers and VMs because anything you fix here is likely to be broken again by the problems in your other thread. You need a reliable basic NAS before you even think about the complicated stuff.

 

Unless, of course each thread is about a different server. (Actually, no, the diagnostics in both threads are for Tower2.)

Sorry. I started this thread once all of my discs were up and running and all seemed well. I did not want to confuse the issue, but I guess both of these are related.

22 minutes ago, levster said:

Sorry. I started this thread once all of my discs were up and running and all seemed well. I did not want to confuse the issue, but I guess both of these are related.

I know there was no bad intention but it had me confused for a while! Then I recognised your rather scary avatar and made the connection. Yes, it looks like your server's dropping cache disks as well. I suggest you leave this thread in limbo for a while and concentrate on the other one, then come back here if necessary.

8 hours ago, levster said:

I looked a little at the scrub but cannot find a guide how to use it.

On the main page click on cache, then there's a scrub button.

7 hours ago, johnnie.black said:

On the main page click on cache, then there's a scrub button.

Great! I'll give it a try. Thank you very much!

On 11/18/2018 at 7:00 PM, levster said:

VM is back! I ran the backup and it seems to have restored fine!

 

Thank you very much!

Not quite sure what you mean by this, but taken at face value it was the worst thing to do.  Running another backup would have backed up the libvirt.img (if you include that in the backup options), and would have over written the good backup with a corrupted image.

 

4 hours ago, Squid said:

Not quite sure what you mean by this, but taken at face value it was the worst thing to do.  Running another backup would have backed up the libvirt.img (if you include that in the backup options), and would have over written the good backup with a corrupted image.

 

What I meant was - I restored the libvirt and dockers from a backup, not run a new backup.

Edited November 20, 20187 yr by levster