Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 30 Dec 2015 01:03:18 +0000
From:      bugzilla-noreply@freebsd.org
To:        freebsd-amd64@FreeBSD.org
Subject:   [Bug 205708] Kernel panic possibly happening while adding/removing/listing ZFS snapshot
Message-ID:  <bug-205708-6@https.bugs.freebsd.org/bugzilla/>

next in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D205708

            Bug ID: 205708
           Summary: Kernel panic possibly happening while
                    adding/removing/listing ZFS snapshot
           Product: Base System
           Version: 10.2-RELEASE
          Hardware: amd64
                OS: Any
            Status: New
          Severity: Affects Only Me
          Priority: ---
         Component: kern
          Assignee: freebsd-bugs@FreeBSD.org
          Reporter: root_dfjo1hpy@meo.ws
                CC: freebsd-amd64@FreeBSD.org
                CC: freebsd-amd64@FreeBSD.org

Hi,

recently my FreeBSD system running ZFS started to crash every now and then.=
 I
think the bug is somehow associated with the amount of datastores*snapshots=
 I'm
having.

The system was actually running fine with snapshots done/deleted every 10
minutes, every 30 minutes, every hour, 6 hours, 12 hours, 24 hours and every
week until it started to behave strange and started to crash.

I then rebooted the system, checked everything and even replaced hardware t=
hat
might have caused this (memory, CPU, motherboard, HBA, basically everything=
 but
the disks itself) with the same or similar hardware that was available.

Then suddenly, the second crash, this time FreeBSD didn't do any kernel dum=
p or
anything similar but just print the stack trace on the monitor and freeze. =
And
stupid I was I rebooted without taking a picture, but I remember reading the
same lines I had in the first crash (which is attached).

Stuff I changed about 1 day before the first crash happened:

sysctl vfs.zfs.l2arc_noprefetch=3D1
zfs set sync=3Ddisabled data
zfs set logbias=3Dthroughput data

The system performed faster after these changes, the main reason was that t=
he
performance was not good enough to keep up with our virtual servers that we=
're
running from iSCSI and NFS4.


This is all I know at this time that should go in this bug report. Attached=
 are
some files from the first crash that occurred.

--=20
You are receiving this mail because:
You are on the CC list for the bug.=



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-205708-6>