Hi. I have an array drive with red X. Background is that cache was out of free space,

I stopped the array but it couldn't be stopped ("retrying to unmount"). I rebooted the server, started back the array, and parity check began to run. I didn't notice any issue and began to copy from cache to Disk 16, and several minutes afterwards it got a red X.

Diagnostics attached. I'm backing up now this disk content to one of the other disks. This will take a lot of time.

I stopped the parity check.

 

Afterwards should I start in maintenance mode and check the file system to check if the emulated FS is ok?

Diagnostics attached.

 

Edit: Now another disk got an X. Attached updated diagnostics.

 

juno-diagnostics-20220212-0219.zip

Edited February 12, 20224 yr by Gico

Looks more like power/connection issues, and both disks look healthy, check/replace cables.

Restarted array in maintenance mode,

repaired disk11 with xfs_repair. Disk16 required xfs_repair -L to be repaird.

Now I restarted the array and still have redX and content is emulated. Disk16 has lost+found directory (expected).

So is it now:

stop array

unassign both disks

start array

stop array

assign disks

start array

1 hour ago, Gico said:

So is it now

Yes, assuming you have fixed your hardware issues.

Thanks. I checked that the cables are connected and so, but this is still quite strange.

In the past I had several times of multiple disks issues at the same time and even loss of data due to FS issues,

until I understood that I need to actively cool the onboard HBA.

Since then it was smooth sailing, and now it's the same symptoms, but fan is working,

the server is not stressed with activity and it's winter so cooler temps, so that's worrying.