I am running zpool 22 (Solaris 10U9) and I am looking for a way to
determine how much more work has to be done to complete a resilver
operation (it is already at 100%, but I know that is not a really
accurate number).
From my understanding of how the resilver operation works, it
walks the metadata structure and then the transaction groups. So if
there is no write (or snapshot or clone or ...) activity, once it
completes the walk of the metadata it is done (I assume the % complete
number is based on this). If there is write activity, it then replays
the TXG that came in after the resilver started. I have two zpools
that are resilvering and are having write activity. I know data is
still being committed to the devices being resilvered, but I am
looking for a way to determine how close they are to being done.
So is there a kernel structure I can look at (with kstat or mdb)
that will tell me how many TXG remain to be written to complete the
resilver ? I know this will be a dynamic number, but it will be a help
to determining if we should idle the replication job (in one of our
two cases) and catch up later (the replication happens over a WAN
link, so it is not very fast, 3 MB/sec. maybe) or just wait it out.
I'll be honest, I am nervous with a raidz2 vdev not at full
strength, and I am looking for some comfort :-)
--
{--------1---------2---------3---------4---------5---------6---------7---------}
Paul Kraus
-> Senior Systems Architect, Garnet River ( http://www.garnetriver.com/ )
-> Sound Coordinator, Schenectady Light Opera Company (
http://www.sloctheater.org/ )
-> Technical Advisor, RPI Players
_______________________________________________
zfs-discuss mailing list
zfs-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
