Debian Codegraph @

upstream/debian-unstable/man8/btrfs-rescue.8

raw
1 .\" Man page generated from reStructuredText
2 .\" by the Docutils 0.22.4 manpage writer.
3 .
4 .
5 .nr rst2man-indent-level 0
6 .
7 .de1 rstReportMargin
8 \\$1 \\n[an-margin]
9 level \\n[rst2man-indent-level]
10 level margin: \\n[rst2man-indent\\n[rst2man-indent-level]]
11 -
12 \\n[rst2man-indent0]
13 \\n[rst2man-indent1]
14 \\n[rst2man-indent2]
15 ..
16 .de1 INDENT
17 .\" .rstReportMargin pre:
18 . RS \\$1
19 . nr rst2man-indent\\n[rst2man-indent-level] \\n[an-margin]
20 . nr rst2man-indent-level +1
21 .\" .rstReportMargin post:
22 ..
23 .de UNINDENT
24 . RE
25 .\" indent \\n[an-margin]
26 .\" old: \\n[rst2man-indent\\n[rst2man-indent-level]]
27 .nr rst2man-indent-level -1
28 .\" new: \\n[rst2man-indent\\n[rst2man-indent-level]]
29 .in \\n[rst2man-indent\\n[rst2man-indent-level]]u
30 ..
31 .TH "BTRFS-RESCUE" "8" "May 10, 2026" "7.0" "BTRFS"
32 .SH NAME
33 btrfs-rescue \- recover a damaged btrfs filesystem
34 .SH SYNOPSIS
35 .sp
36 \fBbtrfs rescue\fP <subcommand> <args>
37 .SH DESCRIPTION
38 .sp
39 A set of commands that are targeting to fix a specific problem and may not
40 suitable for :doc\(gabtrfs\-check\(ga.
41 .sp
42 \fBNote:\fP
43 .INDENT 0.0
44 .INDENT 3.5
45 All the subcommands cannot handle an ongoing replace/balance, please finish/cancel
46 the replace/balance before using this command.
47 .UNINDENT
48 .UNINDENT
49 .SH SUBCOMMAND
50 .INDENT 0.0
51 .TP
52 .B chunk\-recover [options] <device>
53 Recover the chunk tree by scanning the devices
54 .sp
55 \fBOptions\fP
56 .INDENT 7.0
57 .TP
58 .B \-y
59 assume an answer of \fIyes\fP to all questions.
60 .TP
61 .B \-h
62 help.
63 .TP
64 .B \-v
65 (deprecated) alias for global \fI\-v\fP option
66 .UNINDENT
67 .UNINDENT
68 .sp
69 \fBNote:\fP
70 .INDENT 0.0
71 .INDENT 3.5
72 Since \fBchunk\-recover\fP will scan the whole device, it will be very
73 slow especially if executed on a large device.
74 .UNINDENT
75 .UNINDENT
76 .INDENT 0.0
77 .TP
78 .B fix\-device\-size <device>
79 Fix device size and super block total bytes values that do not match.
80 .sp
81 Kernel 4.11 starts to check the device size more strictly and this might
82 mismatch the stored value of total bytes. See the exact error message below.
83 Newer kernel will refuse to mount the filesystem where the values do not match.
84 This error is not fatal and can be fixed. This command will fix the device
85 size values if possible.
86 .INDENT 7.0
87 .INDENT 3.5
88 .sp
89 .EX
90 BTRFS error (device sdb): super_total_bytes 92017859088384 mismatch with fs_devices total_rw_bytes 92017859094528
91 .EE
92 .UNINDENT
93 .UNINDENT
94 .sp
95 The mismatch may also exhibit as a kernel warning:
96 .INDENT 7.0
97 .INDENT 3.5
98 .sp
99 .EX
100 WARNING: CPU: 3 PID: 439 at fs/btrfs/ctree.h:1559 btrfs_update_device+0x1c5/0x1d0 [btrfs]
101 .EE
102 .UNINDENT
103 .UNINDENT
104 .TP
105 .B fix\-data\-checksum <device>
106 Selectively fix data checksum mismatch.
107 .sp
108 There is a long existing problem that if a user space program is doing
109 direct IO and modifies the buffer before the write back finished, it
110 can lead to data checksum mismatches.
111 .sp
112 This problem is known but not fixed until upstream release v6.15
113 (backported to older kernels). So it\(aqs possible to hit false data
114 checksum mismatch for any long running btrfs.
115 .sp
116 In that case this program can be utilized to repair such problem.
117 .sp
118 \fBOptions\fP
119 .INDENT 7.0
120 .TP
121 .B \-r|\-\-readonly
122 readonly mode, only scan for and report data checksum mismatches,
123 do not repair
124 .TP
125 .B \-i|\-\-interactive
126 interactive mode, ask for how to repair, ignore the errors by default
127 .TP
128 .B \-m|\-\-mirror <num>
129 use specified mirror to update the checksum item for all corrupted blocks.
130 .sp
131 The value must be >= 1, and if the corrupted block has fewer mirrors than
132 the value, the mirror number will be \fBnum % (num_mirrors + 1)\fP\&.
133 .UNINDENT
134 .UNINDENT
135 .INDENT 0.0
136 .TP
137 .B clear\-ino\-cache <device>
138 Remove leftover items pertaining to the deprecated \fIinode number cache\fP feature.
139 .sp
140 The feature enabled by mount option \fIinode_cache\fP has been completely
141 removed in 5.11 kernel.
142 .TP
143 .B clear\-space\-cache <v1|v2> <device>
144 Completely remove the on\-disk data of free space cache of given version.
145 .sp
146 Especially for v1 free space cache, \fIclear_cache\fP mount option would only
147 remove the cache for updated block groups, the remaining would not be removed.
148 Thus this command is provided to manually clear the free space cache.
149 .UNINDENT
150 .INDENT 0.0
151 .TP
152 .B clear\-uuid\-tree <device>
153 Clear the UUID tree, so that kernel can regenerate it at next read\-write
154 mount.
155 .sp
156 Since kernel v4.16 there are more sanity check performed, and sometimes
157 non\-critical trees like UUID tree can cause problems and reject the mount.
158 In such case, clearing UUID tree may make the filesystem to be mountable again
159 without much risk as it\(aqs built from other trees.
160 See also mount option rescan_uuid_tree (in btrfs\-man5) \%<btrfs-man5//#\:mount-option-rescan-uuid-tree>\&.
161 .TP
162 .B super\-recover [options] <device>
163 Recover bad superblocks from good copies.
164 .sp
165 \fBOptions\fP
166 .INDENT 7.0
167 .TP
168 .B \-y
169 assume an answer of \fIyes\fP to all questions.
170 .TP
171 .B \-v
172 (deprecated) alias for global \fI\-v\fP option
173 .UNINDENT
174 .TP
175 .B zero\-log <device>
176 Clear the filesystem log tree.
177 .sp
178 This command will clear the filesystem log tree. This may fix a specific
179 set of problem when the filesystem mount fails during log replay. See below
180 for sample stack traces that may show up in system log.
181 .sp
182 \fBNote:\fP
183 .INDENT 7.0
184 .INDENT 3.5
185 Clearing the log may lead to loss of changes that were made
186 since the last transaction commit. This may be up to 30 seconds
187 (default commit period) or less if the commit was implied by
188 other filesystem activity.
189 .UNINDENT
190 .UNINDENT
191 .sp
192 One can determine whether \fBzero\-log\fP is needed according to the kernel
193 backtrace:
194 .INDENT 7.0
195 .INDENT 3.5
196 .sp
197 .EX
198 ? replay_one_dir_item+0xb5/0xb5 [btrfs]
199 ? walk_log_tree+0x9c/0x19d [btrfs]
200 ? btrfs_read_fs_root_no_radix+0x169/0x1a1 [btrfs]
201 ? btrfs_recover_log_trees+0x195/0x29c [btrfs]
202 ? replay_one_dir_item+0xb5/0xb5 [btrfs]
203 ? btree_read_extent_buffer_pages+0x76/0xbc [btrfs]
204 ? open_ctree+0xff6/0x132c [btrfs]
205 .EE
206 .UNINDENT
207 .UNINDENT
208 .sp
209 If the errors are like above, then \fBzero\-log\fP should be used to clear
210 the log and the filesystem may be mounted normally again. The keywords to look
211 for are \(aqopen_ctree\(aq which says that it\(aqs during mount and function names
212 that contain \fIreplay\fP, \fIrecover\fP or \fIlog_tree\fP\&.
213 .UNINDENT
214 .SH EXIT STATUS
215 .sp
216 \fBbtrfs rescue\fP returns a zero exit status if it succeeds. Non zero is
217 returned in case of failure.
218 .SH AVAILABILITY
219 .sp
220 \fBbtrfs\fP is part of btrfs\-progs. Please refer to the documentation at
221 \%<https://\:btrfs\:.readthedocs\:.io>\&.
222 .SH SEE ALSO
223 .sp
224 btrfs\-check(8) \%<>,
225 btrfs\-scrub(8) \%<>,
226 mkfs.btrfs(8) \%<>
227 .\" End of generated man page.