Date: Mon, 25 Jan 2010 15:44:19 -0800 (PST) From: Paul Pathiakis <pathiaki2@yahoo.com> To: freebsd-fs@freebsd.org Subject: ZFS - scrub lead to corruption? Message-ID: <419976.64363.qm@web110515.mail.gq1.yahoo.com>
next in thread | raw e-mail | index | archive | help
Hi,
My ZFS-only machine running on 8.0-stable as of ???? is down hard. It is a mirrored pair of Segate 1500 GB drives (yes, I know the firmware history of these drives and they are a late version.) The sequence of events is as follows:
1) cvsup ports
2) portupgrade -a -r -p
3) everything completes successfully.
4) cvsup the latest 8.0 code.
5) make buildworld
6) while it was compiling, it requested m4 and said "not found"
7) which m4 -> not found
8) I find that /usr/bin/m4 is "there" with ls /usr/bin/m4.
9) cd /usr/ports/devel/m4 -> /usr
10) make; make deinstall; make package; no issues.
11) cd /usr/src; make buildworld. Same behavior.
12) ls -sal /usr/bin/m4 -> Panic
13) Reboot
14) zpool status web
everything is fine and online. :-) No errors.
15) zpool scrub web
16) scrub continues to over 75% (I wasn't watching for the exact #) and then *panic* level 12
17) I reboot single user.
18) It mounts the web at /
19) At that point, I'm deciding what to do and .... Panic level 12 (I assume ZFS was scrubbing in the background and killed me.
20) Help!
How do I recover from this and what additional information does anyone need? I know there's no fsck on ZFS but how do I rollback prior to the writes that caused this? Also, I never enabled snapshots or anything else outside of mirroring.
Thank you for the help,
Paul
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?419976.64363.qm>