zfs discuss - Feb 2007 - unable to mount legacy vol - panic in zfs:space_map_remove - zdb crashes

I have a 100gb SAN lun in a pool, been running ok for about 6 months. panicked
the system this morning. system was running S10U2. In the course of
troubleshooting I''ve installed the latest recommended bundle including
kjp 118833-36 and zfs patch 124204-03

created as:
zpool create zfspool01 /dev/dsk/emcpower0c
zfs create zfspool01/nb60openv
zfs set mountpoint=legacy zfspool01/nb60openv
mkdir -p /zfs/NB60/nb60openv
mount -F zfs zfspool01/nb60openv /zfs/NB60/nb60openv


The mount command now causes a panic:
zfs: WARNING: ZFS replay transaction error 5, dataset zfspool01/nb60openv, seq
0x4180eb0, txtype 9

panic[cpu1]/thread=2a100b75cc0: assertion failed: ss != NULL, file:
../../common/fs/zfs/space_map.c, line: 125

000002a100b74c40 genunix:assfail+74 (7b252450, 7b252460, 7d, 183d400, 11eb000,
0)
  %l0-3: 0000000000000000 0000000000000000 00000000011e5368 000003000b6d2528
  %l4-7: 00000000011eb000 0000000000000000 000000000186f800 0000000000000000
000002a100b74cf0 zfs:space_map_remove+b8 (60001db9eb8, 17698c0000, 20000,
7b252400, 7b252400, 7b252400)
  %l0-3: 0000000000000000 00000017698e0000 00000017623a0000 000003000b6d4fd8
  %l4-7: 000003000b6d5050 0000001762360000 000000007b252000 00000017623e0000
...

Noticing the lun was nearly full, I added a 2nd 100g lun to the pool.

Multiple runs of ''zpool scrub'' have not fixed anything...

sjcbackup01# zpool status -xv
  pool: zfspool01
 state: ONLINE
status: One or more devices has experienced an error resulting in data
        corruption.  Applications may be affected.
action: Restore the file in question if possible.  Otherwise restore the
        entire pool from backup.
   see: http://www.sun.com/msg/ZFS-8000-8A
 scrub: none requested
config:

        NAME          STATE     READ WRITE CKSUM
        zfspool01     ONLINE       0     0     0
          emcpower0c  ONLINE       0     0     0
          emcpower1c  ONLINE       0     0     0

errors: The following persistent errors have been detected:

          DATASET              OBJECT  RANGE
          zfspool01/nb60openv  292     1835008-1966080
          zfspool01/nb60openv  292     2752512-2883584

I can get the info about the object:

sjcbackup01# zdb -vvv zfspool01/nb60openv 292
Dataset zfspool01/nb60openv [ZPL], ID 20, cr_txg 23, last_txg 2158772, 92.3G,
59005 objects, rootbp [L0 DMU objset] 400L/200P DVA[0]=<0:18c0349600:200>
DVA[1]=<0:6c0329600:200> DVA[2]=<0:102040a200:200> fletcher4 lzjb BE
contiguous birth=2158772 fill=59005
cksum=103d9a0f86:66dfb432a03:14f23e0662778:2ebcc4bd282dfb

    ZIL header: claim_txg 2154406, seq 0

        first block: [L0 ZIL intent log] 13000L/13000P
DVA[0]=<0:7208e9000:13000> zilog uncompressed BE contiguous birth=2154403
fill=0 cksum=5f8aab176130501f:6d23a2551b6eb2cf:14:121ac8
...
   Object  lvl   iblk   dblk  lsize  asize  type
       292    3    16K   128K  25.1M  25.1M  ZFS plain file
                                 264  bonus  ZFS znode
        path    /db/data/EMM_DATA.db
        atime   Sat Feb 10 08:44:07 2007
        mtime   Sat Feb 10 09:40:19 2007
        ctime   Sat Feb 10 09:40:19 2007
        crtime  Sun Oct  8 16:15:31 2006
        gen     150
        mode    100600
        size    26226688
        parent  288
        links   1
        xattr   0
        rdev    0x0000000000000000





but zdb crashes:

sjcbackup01# zdb -c zfspool01

Traversing all blocks to verify checksums and verify nothing leaked ...
Assertion failed: ss == NULL, file ../../../uts/common/fs/zfs/space_map.c, line
81, function space_map_add
Abort (core dumped)

I take it there''s some problem replaying the transaction log, but could
use some hints...

Thanks.
 
 
This message posted from opensolaris.org