1BTRFS-SCRUB(8) BTRFS BTRFS-SCRUB(8)
2
6 btrfs-scrub - scrub btrfs filesystem, verify block checksums
7
9 btrfs scrub <subcommand> <args>
10
12 Scrub is a pass over all filesystem data and metadata and verifying the
13 checksums. If a valid copy is available (replicated block group pro‐
14 files) then the damaged one is repaired. All copies of the replicated
15 profiles are validated.
16 NOTE:
18 Scrub is not a filesystem checker (fsck) and does not verify nor re‐
19 pair structural damage in the filesystem. It really only checks
20 checksums of data and tree blocks, it doesn't ensure the content of
21 tree blocks is valid and consistent. There's some validation per‐
22 formed when metadata blocks are read from disk but it's not exten‐
23 sive and cannot substitute full btrfs check run.
24 The user is supposed to run it manually or via a periodic system ser‐
26 vice. The recommended period is a month but could be less. The esti‐
27 mated device bandwidth utilization is about 80% on an idle filesystem.
28 The IO priority class is by default idle so background scrub should not
29 significantly interfere with normal filesystem operation. The IO sched‐
30 uler set for the device(s) might not support the priority classes
31 though.
32 The scrubbing status is recorded in /var/lib/btrfs/ in textual files
34 named scrub.status.UUID for a filesystem identified by the given UUID.
35 (Progress state is communicated through a named pipe in file
36 scrub.progress.UUID in the same directory.) The status file is updated
37 every 5 seconds. A resumed scrub will continue from the last saved po‐
38 sition.
39 Scrub can be started only on a mounted filesystem, though it's possible
41 to scrub only a selected device. See btrfs scrub start for more.
42
44 cancel <path>|<device>
45 If a scrub is running on the filesystem identified by path or
46 device, cancel it.
47 If a device is specified, the corresponding filesystem is found
49 and btrfs scrub cancel behaves as if it was called on that
50 filesystem. The progress is saved in the status file so btrfs
51 scrub resume can continue from the last position.
52 resume [-BdqrR] <path>|<device>
54 Resume a cancelled or interrupted scrub on the filesystem iden‐
55 tified by path or on a given device. The starting point is read
56 from the status file if it exists.
57 This does not start a new scrub if the last scrub finished suc‐
59 cessfully.
60 Options
62 see scrub start.
64 start [-BdrRf] <path>|<device>
66 Start a scrub on all devices of the mounted filesystem identi‐
67 fied by path or on a single device. If a scrub is already run‐
68 ning, the new one will not start. A device of an unmounted
69 filesystem cannot be scrubbed this way.
70 Without options, scrub is started as a background process. The
72 automatic repairs of damaged copies are performed by default for
73 block group profiles with redundancy. No-repair can be enabled
74 by option -r.
75 Options
77 -B do not background and print scrub statistics when fin‐
79 ished
80 -d print separate statistics for each device of the filesys‐
82 tem (-B only) at the end
83 -r run in read-only mode, do not attempt to correct any‐
85 thing, can be run on a read-only filesystem
86 -R raw print mode, print full data instead of summary
88 -f force starting new scrub even if a scrub is already run‐
90 ning, this can useful when scrub status file is damaged
91 and reports a running scrub although it is not, but
92 should not normally be necessary
93 Deprecated options
95 -c <ioprio_class>
97 set IO priority class (see ionice(1) manpage) if the IO
98 scheduler configured for the device supports ionice. This
99 is not supported byg BFQ or Kyber but is not supported by
100 mq-deadline.
101 -n <ioprio_classdata>
103 set IO priority classdata (see ionice(1) manpage)
104 -q (deprecated) alias for global -q option
106 status [options] <path>|<device>
108 Show status of a running scrub for the filesystem identified
109 by path or for the specified device.
110 If no scrub is running, show statistics of the last finished
112 or cancelled scrub for that filesystem or device.
113 Options
115 -d print separate statistics for each device of the
117 filesystem
118 -R print all raw statistics without postprocessing as re‐
120 turned by the status ioctl
121 --raw print all numbers raw values in bytes without the B
123 suffix
124 --human-readable
126 print human friendly numbers, base 1024, this is the
127 default
128 --iec select the 1024 base for the following options, ac‐
130 cording to the IEC standard
131 --si select the 1000 base for the following options, ac‐
133 cording to the SI standard
134 --kbytes
136 show sizes in KiB, or kB with --si
137 --mbytes
139 show sizes in MiB, or MB with --si
140 --gbytes
142 show sizes in GiB, or GB with --si
143 --tbytes
145 show sizes in TiB, or TB with --si
146 A status on a filesystem without any error looks like the
148 following:
149 # btrfs scrub start /
151 # btrfs scrub status /
152 UUID: 76fac721-2294-4f89-a1af-620cde7a1980
153 Scrub started: Wed Apr 10 12:34:56 2023
154 Status: running
155 Duration: 0:00:05
156 Time left: 0:00:05
157 ETA: Wed Apr 10 12:35:01 2023
158 Total to scrub: 28.32GiB
159 Bytes scrubbed: 13.76GiB (48.59%)
160 Rate: 2.75GiB/s
161 Error summary: no errors found
162 With some errors found:
164 Error summary: csum=72
166 Corrected: 2
167 Uncorrectable: 72
168 Unverified: 0
169 • Corrected -- number of bad blocks that were repaired from an‐
171 other copy
172 • Uncorrectable -- errors detected at read time but not possible
174 to repair from other copy
175 • Unverified -- transient errors, first read failed but a retry
177 succeeded, may be affected by lower layers that group or split
178 IO requests
179 • Error summary -- followed by a more detailed list of errors
181 found
182 • csum -- checksum mismatch
184 • super -- super block errors, unless the error is fixed imme‐
186 diately, the next commit will overwrite superblock
187 • verify -- metadata block header errors
189 • read -- blocks can't be read due to IO errors
191
193 btrfs scrub returns a zero exit status if it succeeds. Non zero is re‐
194 turned in case of failure:
195 1 scrub couldn't be performed
197 2 there is nothing to resume
199 3 scrub found uncorrectable errors
201
203 btrfs is part of btrfs-progs. Please refer to the documentation at
204 https://btrfs.readthedocs.io.
205
207 mkfs.btrfs(8)
2086.6.2 Nov 24, 2023 BTRFS-SCRUB(8)