Big thanks to John Groenveld for the expert clues. My workstation is back running like normal and I've learned a few things, mostly stuff that I should have known or should have anticipated about ZFS.
>Once your scrub completes, check if the pool is full. That was the problem. It was so close to full that performance was severely degraded. >If it is, maybe activate an earlier BE, boot it milestone=none and >then destroy the errant, latest BE. While I was booted into milestone=none of the BE I thought was bad (it probably wasn't) and the scrub was running, I did some cleanup to free up space. After I had freed up a few large files (a couple of VirtualBox VMs) performance started to be much better. All of this is because I was looking at space available on each mounted filesystem only using traditional "df -k" and friends, when I should have also been looking at "zfs list" and checking both the "AVAIL" column AND the REFER column for the boot env snapshots. That would have given me a better clue that I needed to do some cleanup. The poking around on my system and the reading I've done this weekend have left me with several additional questions, but all of those are best asked separately. I'll follow up in other emails with those. Thanks again to John for tips! Tim _______________________________________________ openindiana-discuss mailing list [email protected] https://openindiana.org/mailman/listinfo/openindiana-discuss
