[Drbd-dev] DRBD8: incomplete resync with extent not found, inactive extent errors

[Montrose, Ernest]

Hi all,
I have been chasing this issue that we have been seeing a lot with not
much success and unfortunately I have a few other drbd issues needing
investigation.  Essentially, resyncing seems to have problem completing
because the activity log and resync cache cannot be cleared successfully
for a long time at the end of resync session.  Here are some logs:
 
On one node ======:
Apr  4 02:24:29 ellwood kernel: drbd1: Began resync as PausedSyncT (will
sync 13336132 KB [3334033 bits set]).
Apr  4 02:24:29 ellwood kernel: drbd1: Writing meta data super block
now.
Apr  4 02:32:52 ellwood kernel: drbd1: aftr_isp( 1 -> 0 )
Apr  4 02:32:52 ellwood kernel: drbd1: conn( PausedSyncT -> SyncTarget )
peer_isp( 1 -> 0 )
Apr  4 02:32:52 ellwood kernel: drbd1: Syncer continues.
Apr  4 02:34:04 ellwood kernel: drbd1: al_complete_io() called on
inactive extent 1532713819
Apr  4 02:39:06 ellwood kernel: drbd1: drbd_rs_complete_io() called, but
extent not found
Apr  4 02:39:06 ellwood kernel: drbd1: al_complete_io() called on
inactive extent 1532713819
Apr  4 02:40:37 ellwood kernel: drbd1: Retrying drbd_rs_del_all() later.
refcnt=1
Apr  4 02:41:07 ellwood kernel: drbd1: Retrying drbd_rs_del_all() later.
refcnt=1
Apr  4 02:41:07 ellwood kernel: drbd1: Retrying drbd_rs_del_all() later.
refcnt=1
Apr  4 02:41:07 ellwood kernel: drbd1: Retrying drbd_rs_del_all() later.
refcnt=1
Apr  4 02:41:08 ellwood kernel: drbd1: Retrying drbd_rs_del_all() later.
refcnt=1
Apr  4 02:41:08 ellwood kernel: drbd1: Retrying drbd_rs_del_all() later.
refcnt=1
Apr  4 02:41:08 ellwood kernel: drbd1: Retrying drbd_rs_del_all() later.
refcnt=1
Apr  4 02:41:09 ellwood kernel: drbd1: Retrying drbd_rs_del_all() later.
refcnt=1
Apr  4 02:41:09 ellwood kernel: drbd1: Retrying drbd_rs_del_all() later.
refcnt=1
Apr  4 02:41:31 ellwood kernel: drbd1: Retrying drbd_rs_del_all() later.
refcnt=1
Apr  4 02:41:31 ellwood kernel: drbd1: Retrying drbd_rs_del_all() later.
refcnt=1
Apr  4 02:41:31 ellwood kernel: drbd1: Retrying drbd_rs_del_all() later.
refcnt=1

On the other node====
Apr  4 02:24:47 jake kernel: drbd1: Began resync as PausedSyncS (will
sync 13336132 KB [3334033 bits set]).
Apr  4 02:24:47 jake kernel: drbd1: Writing meta data super block now.
Apr  4 02:32:52 jake kernel: drbd1: peer_isp( 1 -> 0 )
Apr  4 02:32:52 jake kernel: drbd1: conn( PausedSyncS -> SyncSource )
aftr_isp( 1 -> 0 )
Apr  4 02:32:52 jake kernel: drbd1: Syncer continues.
Apr  4 02:40:37 jake kernel: drbd1: Resync done (total 971 sec; paused
506 sec; 28676 K/sec)
Apr  4 02:40:37 jake kernel: drbd1:  uuid[History_start] now
356284033FC4539B
Apr  4 02:40:37 jake kernel: drbd1:  uuid[Bitmap] now 0000000000000000
Apr  4 02:40:37 jake kernel: drbd1: conn( SyncSource -> Connected )
pdsk( Inconsistent -> UpToDate )
Apr  4 02:40:37 jake kernel: drbd1: Writing meta data super block now.
 
EM--
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.linbit.com/pipermail/drbd-dev/attachments/20070410/1d73e682/attachment.html