I managed to recover my data after 3 days fighting. Few system changes: - disable ZIL - enable readonly mode - disable zil_replay during mount - change function that chooses uberblock
On snv_78 #mdb -kw > zil_disable/W 1 zil_disable: 0 = 0x1 > spa_mode/W 1 spa_mode: 0x3 = 0x1 > vdev_uberblock_compare+0x49/W 1 vdev_uberblock_compare+0x49: 0xffffffff = 0x1 > vdev_uberblock_compare+0x3b/W 1 vdev_uberblock_compare+0x3b: 0xffffffff = 0x1 > zfsvfs_setup+0x60/v 0xeb zfsvfs_setup+0x60: 0x74 = 0xeb These changes allowed me to import zpool and now I'm coping data. After disk failure matrix controller have cleared caches and some data was lost. ZFS is Copy-on-Write system, so all I had to do to recover data is change uberblock ( first ZFS block ). I think ZFS should be able to change uberblock when zpool import fails. This message posted from opensolaris.org _______________________________________________ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
