Date: Tue, 12 Oct 2010 18:35:04 +0200 From: Willem Jan Withagen <wjw@digiware.nl> To: Pawel Jakub Dawidek <pjd@FreeBSD.org> Cc: fs@freebsd.org Subject: Re: ZFS freeze/livelock Message-ID: <4CB48E38.3080409@digiware.nl> In-Reply-To: <20101012153654.GC2197@garage.freebsd.pl> References: <4CB1DD0F.6000209@digiware.nl> <20101012153654.GC2197@garage.freebsd.pl>
next in thread | previous in thread | raw e-mail | index | archive | help
On 2010-10-12 17:36, Pawel Jakub Dawidek wrote: > On Sun, Oct 10, 2010 at 05:34:39PM +0200, Willem Jan Withagen wrote: >> Hi, >> >> Just had my FreeBSD freeze on me with what I would think is sort of an >> livelock.... >> >> While I was receiving zfs snapshots on my data pool. >> >> Top and systat just kept running, >> but anything getting near a shell (and perhaps disk-io) ended up in: >> >> root@zfs.digiware.nl# gpart create -s gpt da6 >> load: 0.00 cmd: csh 12393 [zfsvfs->z_teardown_inactive_lock] 26.12r >> 0.00u 0.00s 0% 2480k >> load: 0.10 cmd: csh 12393 [zfsvfs->z_teardown_inactive_lock] 96.01r >> 0.00u 0.00s 0% 2480k >> >> Trying to execute to execute shutdown -r now had no effect what so ever. >> Neither did the three-finger salute. >> (Well at least not in 60 sec I was willing to wait.) >> >> Only way out of this situation was hard-reset. And I do have to admit I >> like ZFS for the speed it recovers after unexpected reboot. >> >> To bad there was no alt-ctrl-backspace escape to debugger compiled in. >> I'll do that with the next kernel, just in case. >> >> So the only data point I can give is the ^T output above. > > Maybe you still be able to provide backtraces for all processes with > 'procstat -kk'? > > It looks like a deadlock related to 'zfs recv' or maybe unmounting? System has long rebooted..... I was no longer ableto start any new programs. Probably due to pwdbeing on the locked volume. But I'll this in mind for future time. And yes I was recv-ing snapshots from my other/working zfs-system. Having played with that quite some time, I really like that feature. Although it is not yet robust enough. Perhaps due to the above type problem. I'm now running it with all debug/witness flags on, but boy does that make it slow... --WjW
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4CB48E38.3080409>