Good day great peoples,

this day i came to my mashine it was again that all dockerts were down and I would like to get help  to read and understand what happend.

 

I addet the Diagnostics.

 

On my router i have enabled Port forwarding to my mashine with the ports 80 and 433 for cloudflare and i have for my game server enshroudet port 15636-15638 open.

 

I have Cloudflare installed learning reverse proxies last days and it works fine with stuff going over TCP but game servers need UDP sometimes and therefor the enshroudet port was directly open.

My Ngix webserver side or my searching container worked well over cloudflare tunnel

 

So can somebody tell me what this night happend and show me how to read this errors ?

 

Thx for help guys and Girls and Thams

 

tronshome-diagnostics-20240221-1356.zip

After restarting the server and starting the array it seems all dockers are destroyed ?

 

only the Ngix Manager seems as it woudl work 

It looks as though your cache pool has failed to mount as diagnostics contains:

Feb 15 15:34:24 TronsHome emhttpd: mounting /mnt/cache
Feb 15 15:34:24 TronsHome emhttpd: shcmd (736): mkdir -p /mnt/cache
Feb 15 15:34:24 TronsHome emhttpd: /sbin/btrfs filesystem show /dev/sdb1 2>&1
Feb 15 15:34:24 TronsHome emhttpd: ERROR: no btrfs on /dev/sdb1
Feb 15 15:34:24 TronsHome emhttpd: /usr/sbin/zpool import -d /dev/sdb1 2>&1
Feb 15 15:34:24 TronsHome emhttpd: no pools available to import
Feb 15 15:34:24 TronsHome emhttpd: cache: no uuid
Feb 15 15:34:24 TronsHome emhttpd: /mnt/cache mount error: Unsupported or no file system
Feb 15 15:34:24 TronsHome emhttpd: shcmd (737): rmdir /mnt/cache

 

 

i see now in connect tha failure no 137

 

after google serach i found on a forum how to fix code 137 (Exit Code 137 errors happen when a container or pod was terminated due to high memory usage. Your container or Kubernetes pod will be stopped to prevent the excessive memory consumption from affecting your host’s reliability.)

 

12 minutes ago, itimpi said:

It looks as though your cache pool has failed to mount as diagnostics contains:

Feb 15 15:34:24 TronsHome emhttpd: mounting /mnt/cache
Feb 15 15:34:24 TronsHome emhttpd: shcmd (736): mkdir -p /mnt/cache
Feb 15 15:34:24 TronsHome emhttpd: /sbin/btrfs filesystem show /dev/sdb1 2>&1
Feb 15 15:34:24 TronsHome emhttpd: ERROR: no btrfs on /dev/sdb1
Feb 15 15:34:24 TronsHome emhttpd: /usr/sbin/zpool import -d /dev/sdb1 2>&1
Feb 15 15:34:24 TronsHome emhttpd: no pools available to import
Feb 15 15:34:24 TronsHome emhttpd: cache: no uuid
Feb 15 15:34:24 TronsHome emhttpd: /mnt/cache mount error: Unsupported or no file system
Feb 15 15:34:24 TronsHome emhttpd: shcmd (737): rmdir /mnt/cache

 

 

This are my cache Drives

 

can you tell me whicht made teh faulire so i can take her out ?

 

or is it a driver problem ?

 

Sometimes when i log in my Unraid i dont see any shares any VMs and no Docker until few seconds and then they start to appear.

I think you are l;opoking teh wrong failuer yesterday night was 20.Feb on 21 Feb here in Europe so it sould be on that date the failure that appeared 

Also i dont understand why docker uses so mutch Memory in my server even now when all dockers dont work

I can see that the cache drive is not mounted, and that was the first point at which I saw an issue with the cache drive.

1 minute ago, Olivilo said:

Also i dont understand why docker uses so mutch Memory in my server even now when all dockers dont work

If you have anything configured to use /mnt/cache, then since the cache drive has not mounted this would be writing to RAM rather than persistent storage (and thus lost after a reboot).

1 minute ago, itimpi said:

I can see that the cache drive is not mounted, and that was the first point at which I saw an issue with the cache drive.

you mean this ?

 

I thought as cahs they are allways not mountable so my cache drives were not formatted in teh beginning or what dfid i do wrong that they are unmountable?

 

2 minutes ago, itimpi said:

If you have anything configured to use /mnt/cache, then since the cache drive has not mounted this would be writing to RAM rather than persistent storage (and thus lost after a reboot).

so when i undersatand you 

inicialy my cash drives were not addet and were not used, when i installed the dockers teh server had the soft only in Ram and after a failuer reboot they are gone ? Did i get you right?

 

Ok so can i workaround and tell docker to use now my array untiul my cache problem is solved or can we slove the cache probloem fast and i jsut install teh dockkers new?

 

Is my configuration ok?

 

So if i understadn right i have to format the cache drives (SSDs) i got as Pool so they can work Proper.

 

If i Format with this choise it will only format the SSDs and not tha HDD with the Files on it ?

 

When i Start the formatting process do i have to choose a special file system to work as cache propper?

 

 

Yes they need formatting.     Unraid shows which drives will be formatted when you select this option so you should always check that is just the drives you expect.   In your screenshot it shows the two SSDs will be formatted.

Ok after first formatting it looks like this

first got a file systrem but it doint format both and both are not useabel still

after second formating it keeps same picture

i do neo w exended self test on both discs

and will download teh smart reports and upload here afterwards

Just now, Olivilo said:

will download teh smart reports and upload here afterwards

They are included in the diagnostics so simply uploading a new set of those is better.

ok i run diagnostics again it seems that smart only tells that the drives dont work

tronshome-diagnostics-20240221-1623.zip tronshome-smart-20240221-0706.zip tronshome-smart-20240221-0654.zip

Ok i changed the one SSD that seemed to have errors and splited every ssd in a own cache pool

Now i need only to find out how to move docker on to cache again so it does not use Ram

i would liek to have docker on this 300Gig ssd and the VMs on the 500Gig SDD

Don't know why OP marked this as solved then started another thread about it. I have locked this one. Go here for more: