other: 0.136
vnc: 0.107
files: 0.093
KVM: 0.073
debug: 0.071
device: 0.068
graphic: 0.067
PID: 0.067
semantic: 0.062
socket: 0.059
performance: 0.057
network: 0.056
permissions: 0.050
boot: 0.032
debug: 0.266
files: 0.199
KVM: 0.184
other: 0.057
device: 0.052
PID: 0.047
performance: 0.047
semantic: 0.034
vnc: 0.026
boot: 0.020
network: 0.019
graphic: 0.018
permissions: 0.017
socket: 0.014

Disk corrupted after snapshot deletion

  I found the vm disk corruption after snapshot deletion sometimes(the probability is very low, I'm afraid i can't reproduce it). And I found there is a patch for it as follow, but I'm not sure whether the patch repaired the bug. 
  Drain disk before snapshot deletion can't guarantee anything, there is still pending IO in snapshot-deletion process. Anyone can help?

author	Zhang Haoyu <email address hidden>	2014-10-21 16:38:01 +0800
committer	Stefan Hajnoczi <email address hidden>	2014-11-03 09:48:42 +0000
commit	3432a1929ee18e08787ce35476abd74f2c93a17c (patch)
tree	13a81c0a46707d91622f1593ccf7b926935371fd /block/snapshot.c
parent	573742a5431a99ceaba6968ae269cee247727cce (diff)
snapshot: add bdrv_drain_all() to bdrv_snapshot_delete() to avoid concurrency problem
If there are still pending i/o while deleting snapshot,
because deleting snapshot is done in non-coroutine context, and
the pending i/o read/write (bdrv_co_do_rw) is done in coroutine context,
so it's possible to cause concurrency problem between above two operations.
Add bdrv_drain_all() to bdrv_snapshot_delete() to avoid this problem.

Signed-off-by: Zhang Haoyu <email address hidden>
Reviewed-by: Paolo Bonzini <email address hidden>
Message-id: <email address hidden>
Signed-off-by: Stefan Hajnoczi <email address hidden>
Diffstat (limited to 'block/snapshot.c')
-rw-r--r--	block/snapshot.c	4	
1 files changed, 4 insertions, 0 deletions
diff --git a/block/snapshot.c b/block/snapshot.c
index 85c52ff..698e1a1 100644
--- a/block/snapshot.c
+++ b/block/snapshot.c
@@ -236,6 +236,10 @@ int bdrv_snapshot_delete(BlockDriverState *bs,
         error_setg(errp, "snapshot_id and name are both NULL");
         return -EINVAL;
     }
+
+    /* drain all pending i/o before deleting snapshot */
+    bdrv_drain_all();
+
     if (drv->bdrv_snapshot_delete) {
         return drv->bdrv_snapshot_delete(bs, snapshot_id, name, errp);
     }

My qemu version is 2.1.2

Please try again with the latest release of QEMU first (currently 2.9, or the latest RC of 2.10). QEMU 2.1.2 is very old, the bug might have been fixed in the latest release already.

Thanks for your reply
For some reason, I can't update my version. I believe the latest version have fixed the problem in commit 3432a1929ee18e08787ce35476abd74f2c93a17c or 27a7649a48f9019fa5bd2998d8e342791397bdda.  But I'm not sure the first patch have fixed the problem, as the second patch is too much change for me.

[Expired for QEMU because there has been no activity for 60 days.]