zfs panic, solaris_assert

Damien Fleuriot ml at my.gd
Sat Dec 15 21:05:20 UTC 2012


On 15 December 2012 15:36, Attila Nagy <bra at fsn.hu> wrote:
> On 12/15/2012 03:19 PM, Attila Nagy wrote:
>>
>> Hi,
>>
>> Since running svn revision r243704 I get frequent panics:
>> panic: solaris assert: sa.sa_magic == 0x2F505A (0x4f22a8ed == 0x2f505a),
>> file:
>> /data/usr/src/sys/modules/zfs/../../cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_vfsops.c,
>> line: 597
>> cpuid = 0
>> KDB: stack backtrace:
>> db_trace_self_wrapper() at db_trace_self_wrapper+0x2a
>> kdb_backtrace() at kdb_backtrace+0x37
>> panic() at panic+0x1ce
>> assfail3() at assfail3+0x29
>> zfs_space_delta_cb() at zfs_space_delta_cb+0xbe
>> dmu_objset_userquota_get_ids() at dmu_objset_userquota_get_ids+0x142
>> dnode_sync() at dnode_sync+0xc5
>> dmu_objset_sync_dnodes() at dmu_objset_sync_dnodes+0x5d
>> dmu_objset_sync() at dmu_objset_sync+0x17f
>> dsl_pool_sync() at dsl_pool_sync+0xca
>> spa_sync() at spa_sync+0x34a
>> txg_sync_thread() at txg_sync_thread+0x139
>> fork_exit() at fork_exit+0x11f
>> fork_trampoline() at fork_trampoline+0xe
>> --- trap 0, rip = 0, rsp = 0xffffff90231accf0, rbp = 0 ---
>>
>> I can't tell whether it's the data or the code. If the latter, is this
>> fixed in later revisions?
>> If it's the file system, what can I do with this?
>>
> It seems this is introduced with the following mega-MFC:
> r243674 | mm | 2012-11-29 15:05:04 +0100 (Thu, 29 Nov 2012) | 223 lines
>

For what it's worth, running on amd64 with 4gb ram:
10-CURRENT r244183: Thu Dec 13 15:35:28 UTC 2012

And not experiencing ZFS/solaris problems.

Loader tunables:
vm.kmem_size="3072M"
vfs.zfs.arc_min="128M"
vfs.zfs.arc_max="2048M"


More information about the freebsd-fs mailing list