1MOUNT.NILFS2(8) System Manager's Manual MOUNT.NILFS2(8)
2
3
4
6 mount.nilfs2 - mount a NILFS2 file system
7
9 mount -t nilfs2 [-finrvw] [-o options [,...]] device dir
10 mount -t nilfs2 [-finrvw] [-o options [,...]] device | dir
11
12 mount.nilfs2 [-fnrvw] [-o options [,...]] device dir
13
14 mount.nilfs2 [-V]
15
17 mount.nilfs2 serves to attach a NILFS2 file system on the specified
18 directory dir. It is intended to be executed from mount(8), and will
19 invoke the garbage collector nilfs_cleanerd(8) after an actual mount
20 system call has succeeded. Conversely, umount.nilfs2(8) will shutdown
21 the garbage collector before detaching the file system.
22
23 The standard command line interface is the first form:
24 mount -t nilfs2 [options] device dir
25 This tells the kernel to attach the NILFS2 file system on device at the
26 directory dir. With the second form, the mount program tries to find
27 out a missing device or dir argument from the /etc/fstab table.
28
29 The third form, which directly invokes mount.nilfs2, is also usable
30 since mount.nilfs2 maintains by itself the system mount state such as
31 the list of mounted file systems described in /etc/mtab. However, the
32 first or the second form is usually recommended because some expansive
33 options are not supported by the third form.
34
36 The full set of options used by an invocation of mount(8) is determined
37 by extracting the options from the fstab table, then applying any
38 options specified by the -o argument, and finally applying a -r or -w
39 option, when present.
40
41 See mount(8) for the full set of options. Commonly used options are as
42 follows:
43
44 -V Output version.
45
46 -f Fakes mounting the file system, meaning that the actual system
47 call will be skipped. This option is used to add entries for
48 devices that were mounted earlier with the -n option. It can
49 also be used for invoking nilfs_cleanerd(8) skipped previously.
50
51 -i Don't call mount.nilfs2. This disables garbage collection and
52 handling of pseudo mount options.
53
54 -n Mount without writing in /etc/mtab. This is necessary for exam‐
55 ple when /etc is on a read-only file system.
56
57 -r Mount the file system read-only. A synonym is "-o ro".
58
59 -v Verbose mode.
60
61 -w Mount the file system read/write. This is the default. A synonym
62 is "-o rw".
63
64 -o Options are specified with a -o flag followed by a comma sepa‐
65 rated string of options. Some of these options are only useful
66 when they appear in the /etc/fstab file. For standard filesys‐
67 tem options, see mount(8).
68
70 The following options apply only to the NILFS2 filesystem. They all
71 follow the -o flag.
72
73 barrier / nobarrier
74 These options enable/disable (default is enabled) barrier writes
75 for the block I/O to a lower device. The barrier write serves
76 an important role to ensure consistency of filesystems after a
77 system crash or power failure. NILFS2 uses this feature by
78 default to assure the reliability. For devices not supporting
79 the barrier write, it will be disabled automatically and a warn‐
80 ing will be logged.
81
82 cp=checkpoint-number
83 Specify the checkpoint-number of the snapshot to be mounted.
84 Checkpoints and snapshots are listed by lscp(1). Only the
85 checkpoints marked as snapshot are mountable with this option.
86 Note that the read-only mount option must be specified together.
87
88 errors=continue / errors=remount-ro / errors=panic
89 Define the behaviour when an error is encountered. (Either
90 ignore errors and just mark the file system erroneous and con‐
91 tinue, or remount the file system read-only, or panic and halt
92 the system.) The default is remount-ro. In earlier kernels
93 than Linux 2.6.35, continue was the default.
94
95 pp=protection-period
96 Specify the protection-period for the cleaner daemon (in sec‐
97 onds). nilfs_cleanerd never deletes recent checkpoints whose
98 elapsed time from its creation is smaller than protection-
99 period.
100
101 nogc Disable garbage collection. The cleaner daemon will not be
102 started. It can be be started manually, but in that case it
103 must also be stopped manually before unmounting.
104
105 order=relaxed / order=strict
106 Specify order semantics for file data. Metadata is always writ‐
107 ten to follow the POSIX semantics about the order of filesystem
108 operations.
109
110 relaxed
111 Apply relaxed order semantics that allows modified data
112 blocks to be written to disk without making a checkpoint
113 if no metadata update is going. This mode is equivalent
114 to the ordered data mode of the ext3 filesystem except
115 for the updates on data blocks still conserve atomicity.
116 This will improve synchronous write performance for over‐
117 writing. This is the default mode.
118
119 strict Apply strict in-order semantics that preserves sequence
120 of all file operations including overwriting of data
121 blocks. That means, it is guaranteed that no overtaking
122 of events occurs in the recovered file system after a
123 crash. Unlike journaling filesystems, NILFS2 does not
124 write a same block twice to disk. So there is no signif‐
125 icant performance degradation in comparison with the
126 relaxed mode except for file overwriting.
127
128 norecovery
129 Disable recovery of the filesystem on mount. This disables
130 every write access on the device for read-only mounts or snap‐
131 shots. This option will fail for r/w mounts on an unclean vol‐
132 ume.
133
134 discard / nodiscard
135 These options enable/disable (default is disabled) the use of
136 discard/TRIM commands. The discard/TRIM commands are sent to
137 the underlying block device when blocks are freed. This is use‐
138 ful for SSD devices and sparse/thinly-provisioned LUNs. (since
139 2.6.34).
140
142 The return codes of mount.nilfs2 conform to those of mount(8); the fol‐
143 lowing codes could be returned (the bits can be ORed):
144
145 0 success
146
147 1 incorrect invocation or permissions
148
149 2 system error (out of memory, cannot fork, no more loop devices)
150
151 4 internal bug
152
153 8 user interrupt
154
155 16 problems writing or locking /etc/mtab
156
157 32 mount failure
158
159 64 some mount succeeded
160
162 mount.nilfs2 is written by Ryusuke Konishi <konishi.ryusuke@gmail.com>
163 for NILFS2, based on the mount program included in the util-linux pack‐
164 age.
165
167 mount.nilfs2 is part of the nilfs-utils package and is available from
168 https://nilfs.sourceforge.io.
169
171 nilfs(8), mount(8), umount.nilfs2(8), nilfs_cleanerd(8), lscp(1).
172
173
174
175nilfs-utils version 2.2 Apr 2014 MOUNT.NILFS2(8)