fsck, ufs_fsck - Check and repair UFS file systems
/usr/sbin/fsck [fs_options] [filesystem...]
The following options are interpreted by fsck: Uses the specified block
number as the super block for the file system. Block 32 is usually an
alternate super block. Converts UFS Version 4 back to UFS Version 3.
When you use the -B option on a UFS Version 4 file system, it will be
marked as a Version 3 file system if there are no files or directories
with more than 32767 hardlinks or 32765 subdirectories. If the file
system is in the old (static table) format, this option converts it to
the new (dynamic table) format. If the file system is in the new for‐
mat, this option converts it to the old format, provided that the old
format can support the file system configuration. In interactive mode,
fsck lists the direction of the conversion and asks if the conversion
should be done. If you answer "no," no further operations are done on
the file system. If the -p option is specified, the direction of the
conversion is listed and the conversion is performed without user
interaction if possible. The -p option should be used if all the file
systems are being converted simultaneously. The format of a file system
can be determined from the first line of output from the dumpfs com‐
mand. Forces fsck to check the root file system, even when the file
system is mounted as writable. Use this option with caution, because
running fsck on a mounted root file system can cause its files to
become out of synchronization with running system data. If the problem
occurs, the fsck program displays a message recommending you reboot the
system. Limits the number of parallel checks to the number specified.
By default, the limit is the number of disks running one process per
disk. If a smaller limit is given, the disks are checked using a round
robin-type schedule, one file system at a time. Uses the mode speci‐
fied in octal as the permission bits to use when creating the
lost+found directory rather than the default 777. In particular, sys‐
tems that do not want to have lost files accessible by all users on the
system should use a more restrictive set of permissions, such as 700.
Answers "no" to all the prompts except for the “CONTINUE?” prompt. The
option does not write to the lost+found file in the file system. If
you do not have write permission on the file system, fsck defaults to
the behavior of the -n option. Causes fsck to unconditionally check
the file system even if the file system's clean byte is set. That is,
a file system is checked even if it has been unmounted cleanly. Nonin‐
teractively corrects the following file system inconsistencies: unref‐
erenced inodes, link counts in inodes that are too large, missing
blocks in the free map, blocks in the free map that are also in files,
and wrong counts in the super-block. Causes more extensive messages to
be displayed during the file system checks (verbose mode). Answers
"yes" to all the prompts. This option should be used with caution
because the integrity of the file system data can be corrupted by
answering "yes" to all the prompts.
The fsck command is a front-end program for the ufs_fsck program, which
checks and repairs UFS file systems. Do not use this command for AdvFS
file systems: instead, see advfs(4).
The fsck program has more consistency checks than its check, dcheck,
fcheck, and icheck predecessors combined.
You must be root to use this command.
If you do not specify a file system in the command line, the fsck com‐
mand checks the file systems in the /etc/fstab file.
With one exception, the fsck command cannot be used on an active file
system. The command checks to determine whether the partition to be
checked, or an overlapping partition, is in use. The exceptional case
occurs when you run the command on an active root file system and spec‐
ify the -f option. Use this option with caution, because it can cause
the data on the running system (the in-memory data) to become unsyn‐
chronized from the file system data (the on-disk data).
The fsck program interactively repairs inconsistent file system condi‐
tions. If the file system is found to be consistent, the number of
files, blocks used, and free blocks are reported. If the file system
is inconsistent, you are prompted before each correction is attempted.
For each corrected inconsistency, one or more lines are displayed,
identifying the file system on which the correction takes place and the
type of correction. After successfully correcting a file system, the
fsck program displays the number of files on that file system, the num‐
ber of used and free blocks, and the percentage of fragmentation.
The default behavior of the fsck program is to interactively check the
following UFS file system inconsistencies:
Blocks claimed by more than one inode or the free map
Blocks claimed by an inode outside the range of the file system
Incorrect link counts
Size checks: directory size not of proper format; partially truncated file
Bad inode format
Blocks not accounted for anywhere
Directory checks: file pointing to unallocated inode; inode number out of range; . (dot) or .. (dot dot) not the first two entries of a directory or having the wrong inode number
Super Block checks: more blocks for inodes than there are in the file system
Bad free block map format
Total free block and/or free inode count incorrect
If you use the -p option, the fsck program noninteractively attempts to
correct specific file system inconsistencies. The corrections are made
only if they can be done safely. The fsck program can noninteractively
correct the following file system inconsistencies:
Link counts in inodes that are too large
Missing blocks in the free map
Blocks in the free map that are also in files
Wrong counts in the super-block.
If fsck encounters any other inconsistencies, it exits with an abnormal
return status and a subsequent reboot will fail.
The system makes sure that only a restricted class of innocuous incon‐
sistencies can occur unless hardware or software failures intervene.
Note that some of the corrective actions can result in a loss of data.
The amount and severity of data lost can be determined from the diag‐
At system boot, fsck-p runs automatically and reads the /etc/fstab
file to determine which file systems to check. Only partitions that are
mounted rw or ro and have a non-zero pass (1 or more) number are
checked. File systems that have a pass number 1 (usually only the root
file system) are checked one at a time. When pass 1 completes, the
remaining pass numbers are processed with one parallel fsck process
running per disk drive in the same pass.
The per disk drive logic is based on the /dev/disk/dsk0a syntax where
different partition letters are treated as being on the same disk
drive. Partitions layered on top of an LSM device may not follow this
naming convention. Where LSM is used, you can use unique pass numbers
in the /etc/fstab file to sequence the fsck checks.
If a QUIT signal is sent, fsck finishes the file system checks and then
exits with an abnormal return status that causes the automatic reboot
to fail. This is useful if you want to finish the file system checks
during an automatic reboot but do not want the machine to come up
multiuser after the checks complete.
If the fsck default program fails (terminates with a non-zero value),
fsck terminates with the exit value. This ensures that the auto-reboot
dependencies, such as those commonly used in the run command script,
continue to function.
If orphaned files or directories (allocated but unreferenced) are
found, you are asked if you want to reconnect the files and directories
that are not empty by putting them in the lost+found directory. The
program indicates whether the file or directory is empty or not empty.
The fsck program reconnects directories that are not empty and assigns
the inode number for the name. If the lost+found directory does not
exist, it is created, and if it has insufficient space, the size is
increased. Empty files and directories are removed unless you specify
the -n option.
The following message indicates that the system is avoiding a potential
panic by skipping the file system check: /dev/disk/dsk2g: skipping
filesystem already mounted (read-write) on '/usr'
The following message indicates that the file system is mounted:
/dev/disk/dsk2g on /usr: Device busy
The following message indicates that the partition you are checking is
open: ERROR: /dev/rdisk/dsk3a or an overlapping partition is open fsck
cannot be run on an active filesystem
The following is an example of an /etc/fstab file that is used in the
examples in this section:
/dev/disk/dsk2a / ufs rw 1 1 /dev/disk/dsk0g /usr ufs
rw 1 2 /dev/disk/dsk2b swap1 ufs sw 0 2 /dev/disk/dsk0b
/public ufs sw 0 2 /dev/disk/dsk2g /var ufs rw 1 2
/dev/disk/dsk3c /usr/users ufs rw 1 2
The following command checks all the file systems in the /etc/fstab
file but makes no corrections: # fsck-n
The following example checks a file system found in the /etc/fstab file
and checks a file system that is not found in the /etc/fstab file: #
fsck /dev/rdisk/dsk3c /dev/rdisk/dsk1a
The following command noninteractively checks the file systems in the
/etc/fstab file: # fsck-p
Specifies the command path. Specifies the command path. Contains the
default list of file systems to check.