Illumos 6370 - ZFS send fails to transmit some holes

6370 ZFS send fails to transmit some holes
Reviewed by: Matthew Ahrens <[email protected]>
Reviewed by: Chris Williamson <[email protected]>
Reviewed by: Stefan Ring <[email protected]>
Reviewed by: Steven Burgess <[email protected]>
Reviewed by: Arne Jansen <[email protected]>
Approved by: Robert Mustacchi <[email protected]>

References:
  https://www.illumos.org/issues/6370
  https://github.com/illumos/illumos-gate/commit/286ef71

In certain circumstances, "zfs send -i" (incremental send) can produce
a stream which will result in incorrect sparse file contents on the
target.

The problem manifests as regions of the received file that should be
sparse (and read a zero-filled) actually contain data from a file that
was deleted (and which happened to share this file's object ID).

Note: this can happen only with filesystems (not zvols, because they do
not free (and thus can not reuse) object IDs).

Note: This can happen only if, since the incremental source (FromSnap),
a file was deleted and then another file was created, and the new file
is sparse (i.e. has areas that were never written to and should be
implicitly zero-filled).

We suspect that this was introduced by 4370 (applies only if hole_birth
feature is enabled), and made worse by 5243 (applies if hole_birth
feature is disabled, and we never send any holes).

The bug is caused by the hole birth feature. When an object is deleted
and replaced, all the holes in the object have birth time zero. However,
zfs send cannot tell that the holes are new since the file was replaced,
so it doesn't send them in an incremental. As a result, you can end up
with invalid data when you receive incremental send streams. As a
short-term fix, we can always send holes with birth time 0 (unless it's
a zvol or a dataset where we can guarantee that no objects have been
reused).

Ported-by: Steven Burgess <[email protected]>
Signed-off-by: Brian Behlendorf <[email protected]>
Closes #4369
Closes #4050

1 files changed, 7 insertions, 1 deletions

diff --git a/module/zfs/dmu_object.c b/module/zfs/dmu_object.c
index 127a6757f..5faecafc7 100644
--- a/module/zfs/dmu_object.c
+++ b/module/zfs/dmu_object.c
@@ -20,7 +20,7 @@
  */
 /*
  * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
- * Copyright (c) 2013, 2014 by Delphix. All rights reserved.
+ * Copyright (c) 2013, 2015 by Delphix. All rights reserved.
  * Copyright 2014 HybridCluster. All rights reserved.
  */
 
@@ -50,6 +50,12 @@ dmu_object_alloc(objset_t *os, dmu_object_type_t ot, int blocksize,
 		 * reasonably sparse (at most 1/4 full).  Look from the
 		 * beginning once, but after that keep looking from here.
 		 * If we can't find one, just keep going from here.
+		 *
+		 * Note that dmu_traverse depends on the behavior that we use
+		 * multiple blocks of the dnode object before going back to
+		 * reuse objects.  Any change to this algorithm should preserve
+		 * that property or find another solution to the issues
+		 * described in traverse_visitbp.
 		 */
 		if (P2PHASE(object, L2_dnode_count) == 0) {
 			uint64_t offset = restarted ? object << DNODE_SHIFT : 0;