I recently added a cache pool to my Server, then upgrade my cache 2 drive to match my cache 1 drive.  WD black nvme 500GB cache 1 /AData nvme 500GB cache 2 to two WD black nvme 500GB cache 1 /500GB cache 2.  Now the cache 2 drives shows no reads or writes being made. 

 

Can someone please advise?  Thanks in advance, sceenshot below.

 

Cache2 is almost certainly not being used, please post the diagnostics: Tools -> Diagnostics

Thanks for the response... attached.

x-server-diagnostics-20190129-2157.zip

Does this mean anything?

 

Jan 29 20:18:40 X-Server root: /dev/nvme1n1:
Jan 29 20:18:40 X-Server root:  issuing standby command
Jan 29 20:18:40 X-Server emhttpd: shcmd (1798): exit status: 25
Jan 29 20:18:45 X-Server emhttpd: req (4): cmdSpinup=cache2&startState=STARTED&csrf_token=****************
Jan 29 20:18:45 X-Server emhttpd: shcmd (1799): /usr/sbin/hdparm -S0 /dev/nvme1n1
Jan 29 20:18:45 X-Server root:  HDIO_DRIVE_CMD(setidle) failed: Inappropriate ioctl for device
Jan 29 20:18:45 X-Server root: 
Jan 29 20:18:45 X-Server root: /dev/nvme1n1:
Jan 29 20:18:45 X-Server root:  setting standby to 0 (off)
Jan 29 20:18:45 X-Server emhttpd: shcmd (1799): exit status: 25

18 minutes ago, drugdoctor said:

Does this mean anything?

Yes, but unrelated, NVMe devices don't spin down.

 

Cache pool is kind of a mess, cache2 is not part of the pool like suspected, but the remaining member has multiple data profiles including DUP, which is very strange, unless you did a manual balance specifying that, though you could balance and re-add the second device these type of issues can sometimes lead to unpredictable behavior, I suggest you backup you cache to the array, you can use this to help, clear both devices with:

 

blkdiscard /dev/nvmeXn1

Then re-format cache the pool and restore the data.

Looks like it worked (writes after balance).  Thanks for the help!