Hi Unraiders,

 

Horrible surprise today. Nas has been running fine for many weeks with no issues, today I come to it and find cpu load at 128 and completely unresponsive. Display is giving lots of BTRFS errors (I only use BTRFS as cache file system so that narrowed it down). After managing to reboot it, the cache drive is showing up as "unmountable - no file system" and a parity chack has automatically been started. My cache drive is a single Samsung EVO 860 Pro.

 

I used CA-Backup plugin for AppData etc so have recent backups I can restore if necessary.

 

Diagnostics is attached, any advice is welcome and much appreciated. Thankyou.

diagnostics-20181212-1741.zip

Best way forward it to backup cache, re-format and restore, if you already have a current backup just format and restore, if not see here for help with the backup:

https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?do=findComment&comment=543490

 

23 hours ago, johnnie.black said:

Best way forward it to backup cache, re-format and restore, if you already have a current backup just format and restore, if not see here for help with the backup:

https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?do=findComment&comment=543490

 

Thanks very much for the help, was able to be back up and running very quickly.

 

A quick question if you dont mind, how could this have happened and how can I avoid it in the future?

4 minutes ago, Ascii227 said:

how could this have happened and how can I avoid it in the future?

Difficult to say, you mentioned btrfs errors but those could be for example docker image related, and not the cache fs, since last boot was unclean I guess you had to force the reboot, so without the pre-reboot logs difficult to say if the problem was already there or was caused by the unclean shutdown itself, which btrfs can't always handle well, especially is it was writing to a disk when plug was pulled.

2 hours ago, johnnie.black said:

since last boot was unclean I guess you had to force the reboot, so without the pre-reboot logs difficult to say if the problem was already there or was caused by the unclean shutdown itself, which btrfs can't always handle well, especially is it was writing to a disk when plug was pulled. 

Unfortunately I could see no option other than to pull the plug. I spent a good while over ssh trying to get things to respond but the shfs and btrfs processes were through the roof on cpu load, and the system wouldnt even respond to kill or reboot commands.

 

Thanks for the info, next time I will try to do the diagnostics dump before the reboot.