1BTRFS-RESCUE(8) BTRFS BTRFS-RESCUE(8)
2
6 btrfs-rescue - recover a damaged btrfs filesystem
7
9 btrfs rescue <subcommand> <args>
10
12 btrfs rescue is used to try to recover a damaged btrfs filesystem.
13
15 chunk-recover [options] <device>
16 Recover the chunk tree by scanning the devices
17 Options
19 -y assume an answer of yes to all questions.
21 -h help.
23 -v (deprecated) alias for global -v option
25 NOTE:
27 Since chunk-recover will scan the whole device, it will be very slow
28 especially executed on a large device.
29 fix-device-size <device>
31 fix device size and super block total bytes values that do not
32 match
33 Kernel 4.11 starts to check the device size more strictly and
35 this might mismatch the stored value of total bytes. See the ex‐
36 act error message below. Newer kernel will refuse to mount the
37 filesystem where the values do not match. This error is not fa‐
38 tal and can be fixed. This command will fix the device size
39 values if possible.
40 BTRFS error (device sdb): super_total_bytes 92017859088384 mismatch with fs_devices total_rw_bytes 92017859094528
42 The mismatch may also exhibit as a kernel warning:
44 WARNING: CPU: 3 PID: 439 at fs/btrfs/ctree.h:1559 btrfs_update_device+0x1c5/0x1d0 [btrfs]
46 clear-ino-cache <device>
48 Remove leftover items pertaining to the deprecated inode cache
49 feature.
50 The inode cache feature (enabled by mount option "inode_cache")
52 has been completely removed in 5.11 kernel.
53 clear-space-cache <v1|v2> <device>
55 Completely remove the on-disk data of free space cache of given
56 version.
57 Especially for v1 free space cache, clear_cache mount option
59 would only remove the cache for updated block groups, the re‐
60 maining would not be removed. Thus this command is provided to
61 manually clear the free space cache.
62 clear-uuid-tree <device>
64 Clear UUID tree, so that kernel can re-generate it at next
65 read-write mount.
66 Since kernel v4.16 there are more sanity check performed, and
68 sometimes non-critical trees like UUID tree can cause problems
69 and reject the mount. In such case, clearing UUID tree may make
70 the filesystem to be mountable again without much risk as it's
71 built from other trees.
72 super-recover [options] <device>
74 Recover bad superblocks from good copies.
75 Options
77 -y assume an answer of yes to all questions.
79 -v (deprecated) alias for global -v option
81 zero-log <device>
83 clear the filesystem log tree
84 This command will clear the filesystem log tree. This may fix a
86 specific set of problem when the filesystem mount fails due to
87 the log replay. See below for sample stack traces that may show
88 up in system log.
89 The common case where this happens was fixed a long time ago, so
91 it is unlikely that you will see this particular problem, but
92 the command is kept around.
93 NOTE:
95 Clearing the log may lead to loss of changes that were made
96 since the last transaction commit. This may be up to 30 sec‐
97 onds (default commit period) or less if the commit was im‐
98 plied by other filesystem activity.
99 One can determine whether zero-log is needed according to the
101 kernel backtrace:
102 ? replay_one_dir_item+0xb5/0xb5 [btrfs]
104 ? walk_log_tree+0x9c/0x19d [btrfs]
105 ? btrfs_read_fs_root_no_radix+0x169/0x1a1 [btrfs]
106 ? btrfs_recover_log_trees+0x195/0x29c [btrfs]
107 ? replay_one_dir_item+0xb5/0xb5 [btrfs]
108 ? btree_read_extent_buffer_pages+0x76/0xbc [btrfs]
109 ? open_ctree+0xff6/0x132c [btrfs]
110 If the errors are like above, then zero-log should be used to
112 clear the log and the filesystem may be mounted normally again.
113 The keywords to look for are 'open_ctree' which says that it's
114 during mount and function names that contain replay, recover or
115 log_tree.
116
118 btrfs rescue returns a zero exit status if it succeeds. Non zero is re‐
119 turned in case of failure.
120
122 btrfs is part of btrfs-progs. Please refer to the documentation at
123 https://btrfs.readthedocs.io.
124
126 btrfs-check(8), btrfs-scrub(8), mkfs.btrfs(8)
1276.6.2 Nov 24, 2023 BTRFS-RESCUE(8)