"Critical Errors" but ZFS status Healthy?

[WKG]

I have alerts on my new FreeNAS setup as follows:
"CRITICAL: Device: /dev/da2 [SAT], 16 Currently unreadable (pending) sectors"
"CRITICAL: Device: /dev/da2 [SAT], Self-Test Log error count increased from 0 to 1"

Yet under Storage>Volumes>ZFS shows "HEALTHY" status. Should I be worried?

Configuration as follows:
Build FreeNAS-9.3-STABLE-201512121950
Platform Intel(R) Xeon(R) CPU X5570 @ 2.93GHz
Memory 49109MB

HP Z800 workstation, dual X5570, 48gb ECC-RAM, AS#460838-002 motherboard, add-on PCIex LSI LSI00303 (9207-4i4e) SATA/SAS HBA, 3x 4TB HGST DeskStar 7200 SATA drives in RaidZ1-0 (have two spares currently)

 

[Mlovelace]

You need to replace da2, it's dying. The freenas docs will tell you how to replace the bad drive.

FYI, raidz1 is rolling the dice with your data, you'd be better off by miles with raidz2 and a single spare, then the raidz1 and two spares you currently have.

 

[BigDave]

Yet under Storage>Volumes>ZFS shows "HEALTHY" status. Should I be worried?

Not if you have a solid backup plan in effect.

With your level of redundancy, you need to keep a close eye on the disk (da2). An indication of
unreadable sectors is usually the beginning of the end.

 

[WKG]

Okay good to know. Is it possible to upgrade the RaidZ1-0 to RaidZ2 in-place? Or I'm assuming I'll need to move all the data off and then recreate the entire Zpool, correct?

 

[Mlovelace]

Okay good to know. Is it possible to upgrade the RaidZ1-0 to RaidZ2 in-place? Or I'm assuming I'll need to move all the data off and then recreate the entire Zpool, correct?

Yep, its nuke and rebuild for the pool. You can't change it once it in place.

 

[BigDave]

I'm assuming I'll need to move all the data off and then recreate the entire Zpool, correct?

Correct.

 

[WKG]

Fair enough. I will start working on that immediately.