[OpenIndiana-discuss] Huge ZFS root pool slowdown - diagnose root cause?

[Nikola M.]

On 13.12.18. 22:14, Lou Picciano wrote:
> The plot thickens, I’m afraid. Since last post, I’ve replaced the drive, and throughput remains molasses-in-January slow…
> period indicated below is more than 24 hours:
>
>   scan: resilver in progress since Wed Dec 12 15:13:11 2018
>         26.9G scanned out of 1.36T at 345K/s, (scan is slow, no estimated time)
>     26.9G resilvered, 1.93% done

ZFS does not depend on hardware.
I target this could also be the hardware problem.
You can just get drives off the machine and connect them to another
machine and see how it boots on a new machine.
Yes, check cables, SATA ports on the controller..
Also check BIOS/firmware settings, if hard disk controller works in AHCI
or ATA mode.

Also see versions of both your zpool and zfs. (zpool upgrade, zfs upgrade)
I have seen weird things with older ZFS pools on newer illumoses.

Also if controller is SAS and disks are SATA and one is using SAS
expander to connect SATA drives, that would work, but it is a NO-NO,
since SAS expanders could be faulty and unable to debug. Just to mention..

Did replacing drive worked, finished?
Also what is the motherboard, hard disk controller, versions of
osnet_incorporation and userland_incorporation in current boot
environment (BE)?
Have you tried booting system from an older BE ? (Targeting that newer
illumos could have newly acquired problem with your hardware).

You also never had problems with data, since ZFS returned 0 errors.
Could it be you have a problem with machine overheating, and working
very slow?
Anyway, good to know what happened after resilvering with a new drive in
the mirror.