My log files just filled up with what looks like errors.

 

Anyone have any idea what it means, kind of looks like filesystem issues.

 

sdc is one of my 2 SSD Cache drives and I tried running SMART self tests on it but when I click start the button just flashes and does nothing.

aeris-diagnostics-20180225-0710.zip

In the last two days I have also gotten the following emailed to me once a day. I am betting it is related:

 

Email Subject: cron for user root /sbin/fstrim -v /mnt/cache | logger &> /dev/null

Email Body: fstrim: /mnt/cache: FITRIM ioctl failed: Input/output error

Cache2 dropped offline, check cables and run a correcting scrub, check that there are no uncorrectable errors.

Now my cache is starting read only after reboots (NOTHING in Unraid tells me this or that anything is wrong, I have to try to write to it via SSH to know this).

 

New diagnostics after the latest reboot.

 

I am guessing either one of the cache drives died (passing extended smart tests but looks like it might have a new unnallocated sector count that was successfully moved) or the btrfs corupted itself.

aeris-diagnostics-20180226-0918.zip

sdc is corrupt for being offlined, you should still be able st mount the cache if you temporarily physically disconnect sdc, if it mounts OK then wipe sdc and re-add to pool.

My whole cache seems to be hosed. BTRFS is doing some serious crunching on something right now after I removed the questionable drive from the cache pool. Will see if that ever comes back up up with a single drive or if I need to rebuild everything (I have backups!). Still don't know if the drive is bad though but suspecting it might be.

14 minutes ago, yippy3000 said:

Still don't know if the drive is bad though but suspecting it might be.

SMART shows some issues, so it might be the SSD, difficult to say for sure because the syslog only begins after the problem starts.