[Solved] Headaches From 1/2 Reused build. Current isssue, BTFS Error on NVME cache drive and Dirty Shutdowns.

[Mr.N3rd]

The Main issue is the BTFS Error. I can't invoke the mover because of the errors and I/O Errors. the System as it sits is unusable. It even Killed my Docker Containers and the service as a whole.
I tried to stop the array and its just giving off the "emhttpd:shcmd 2925 and 2926 error"

I'm really at a loss. The spinning drives were originally from my other build that has had 4-5 years run time so I know its almost time to replace but the NVME is Brand new Samsung 970 Evo 1TB. So it having errors is baffling to me. This was a Fresh config and install of UnRaid on a Brand new USB drive. I have attached the current log, the errors under the Docker Tab and the current Diagnostic. The system is still running as we speak trying to stop the array. Still cant do it after 30 or so minutes.

the other issue is Unclean shutdowns, I know its been discussed alot but I have tried to shut down multiple times, the first time I let it go for 2 days without realizing it was still not powered off. Its headless so Not sure what the log said I just held down the Power button to kill it. Hoping it was in a safe state and restart it later on. I cannot get it to actually have a clean shutdown...


Specs:
Asus x270-E
Intel 7700K
Corsair Vengenex LPX 64G (Brand New)
Adaptec ASR-71605 SAS HBA
Samsung NVME cache
Two Off brand Sata 3 Cache drives (Old and planning to retire after fixing this issue)


All the Spinning drives are SAS and Either Hitatchi or HSGT (Which is the same company) and with in a couple of models apart. and all 4TB drives

Please help me revive my weekend project before I loose my mind!

UnRaid-diagnostics-20221119-1704.zip

Edited December 6, 2022 by Mr.N3rd
solved tag

[JorgeB]

Pool went read-only because it ran out of space, if you want to fully use the pool capacity with different size devices use the single profile instead of raid0, reboot, see if you can free up some space then run a scrub to see if the errors can be fixed.

[JorgeB]

Forgot to mention, there are macvlan call traces, these can make Unraid crash, switch to ipvlan.

[Mr.N3rd]

Thanks for chiming in! So I "converted to Single Mode" ( I put it in quotations because it didn't seem to do anything, watching the log just showed a command that ran with no result - I assume to be expected)

Nov 20 10:46:59 GifServer  ool www[15476]: /usr/local/emhttp/plugins/dynamix/scripts/btrfs_balance 'start' '/mnt/cache' '-dconvert=single,soft -mconvert=raid1,soft'

I have since rebooted successfully. Will post outcome shortly..

 

From what I can see it was set in the Docker Settings to ipvlan not macvlan

[Mr.N3rd]

[Mr.N3rd]

So the system Rebooted and the cache pool is unavailable and requiring format, to be expected from the change from Raid0 to single

 

[JorgeB]

15 hours ago, Mr.N3rd said:

to be expected from the change from Raid0 to single

It's not expected, did you free up some space before converting?

 

15 hours ago, Mr.N3rd said:

From what I can see it was set in the Docker Settings to ipvlan not macvlan

Strange, are you still on v6.11.1 or updated to v6.11.4?

[Mr.N3rd]

Sorry @JorgeB I didn't realize you replied. I did not free up space. In fact I just let it format and had to redo all the docker containers I had, luckily I didn't have any VM's on this server yet.

and Yes I was on 6.11.4 at the time, Now its on 6.11.5 and a host of new issues...