1 .\" Revision 1.0 93/06/3 23:00 chk
5 .TH TUNE2FS 8 "@E2FSPROGS_MONTH@ @E2FSPROGS_YEAR@" "E2fsprogs version @E2FSPROGS_VERSION@"
7 tune2fs \- adjust tunable filesystem parameters on ext2/ext3/ext4 filesystems
26 .I interval-between-checks
37 .I reserved-blocks-percentage
41 .RI [^]mount-options [,...]
45 .I reserved-blocks-count
73 .I last-mounted-directory
77 .RI [^] feature [,...]
90 allows the system administrator to adjust various tunable filesystem
91 parameters on Linux ext2, ext3, or ext4 filesystems. The current values
92 of these options can be displayed by using the
96 program, or by using the
101 .BI \-c " max-mount-counts"
102 Adjust the number of mounts after which the filesystem will be checked by
106 is 0 or \-1, the number of times the filesystem is mounted will be disregarded
111 Staggering the mount-counts at which filesystems are forcibly
112 checked will avoid all filesystems being checked at one time
113 when using journaled filesystems.
115 You should strongly consider the consequences of disabling
116 mount-count-dependent checking entirely. Bad disk drives, cables,
117 memory, and kernel bugs could all corrupt a filesystem without
118 marking the filesystem dirty or in error. If you are using
119 journaling on your filesystem, your filesystem will
121 be marked dirty, so it will not normally be checked. A
122 filesystem error detected by the kernel will still force
123 an fsck on the next reboot, but it may already be too late
124 to prevent data loss at that point.
128 option for time-dependent checking.
130 .BI \-C " mount-count"
131 Set the number of times the filesystem has been mounted.
132 If set to a greater value than the max-mount-counts parameter
137 will check the filesystem at the next reboot.
139 .BI \-e " error-behavior"
140 Change the behavior of the kernel code when errors are detected.
141 In all cases, a filesystem error will cause
143 to check the filesystem on the next boot.
145 can be one of the following:
149 Continue normal execution.
152 Remount filesystem read-only.
155 Cause a kernel panic.
158 .BI \-E " extended-options"
159 Set extended options for the filesystem. Extended options are comma
160 separated, and may take an argument using the equals ('=') sign.
161 The following extended options are supported:
164 .BI stride= stride-size
165 Configure the filesystem for a RAID array with
167 filesystem blocks. This is the number of blocks read or written to disk
168 before moving to next disk. This mostly affects placement of filesystem
169 metadata like bitmaps at
171 time to avoid placing them on a single disk, which can hurt the performance.
172 It may also be used by block allocator.
174 .BI stripe_width= stripe-width
175 Configure the filesystem for a RAID array with
177 filesystem blocks per stripe. This is typically be stride-size * N, where
178 N is the number of data disks in the RAID (e.g. RAID 5 N+1, RAID 6 N+2).
179 This allows the block allocator to prevent read-modify-write of the
180 parity in a RAID stripe if possible when the data is written.
182 .BI hash_alg= hash-alg
183 Set the default hash algorithm used for filesystems with hashed b-tree
184 directories. Valid algorithms accepted are:
190 .BI mount_opts= mount_option_string
191 Set a set of default mount options which will be used when the file
192 system is mounted. Unlike the bitmask-based default mount options which
193 can be specified with the
196 .I mount_option_string
197 is an arbitrary string with a maximum length of 63 bytes, which is
198 stored in the superblock.
200 The ext4 file system driver will first apply
201 the bitmask-based default options, and then parse the
202 .IR mount_option_string ,
203 before parsing the mount options passed from the
207 This superblock setting is only honored in 2.6.35+ kernels;
208 and not at all by the ext2 and ext3 file system drivers.
211 Set a flag in the filesystem superblock indicating that it may be
212 mounted using experimental kernel code, such as the ext4dev filesystem.
215 Clear the test_fs flag, indicating the filesystem should only be mounted
216 using production-level filesystem code.
220 Force the tune2fs operation to complete even in the face of errors. This
221 option is useful when removing the
223 filesystem feature from a filesystem which has
224 an external journal (or is corrupted
225 such that it appears to have an external journal), but that
226 external journal is not available.
229 Removing an external journal from a filesystem which was not cleanly unmounted
230 without first replaying the external journal can result in
231 severe data loss and filesystem corruption.
234 Set the group which can use the reserved filesystem blocks.
237 parameter can be a numerical gid or a group name. If a group name is given,
238 it is converted to a numerical gid before it is stored in the superblock.
240 .B \-i " \fIinterval-between-checks\fR[\fBd\fR|\fBm\fR|\fBw\fR]"
241 Adjust the maximal time between two filesystem checks.
244 will interpret the number
245 .I interval-between-checks
250 as weeks. A value of zero will disable the time-dependent checking.
252 It is strongly recommended that either
254 (mount-count-dependent) or
256 (time-dependent) checking be enabled to force periodic full
258 checking of the filesystem. Failure to do so may lead to filesystem
259 corruption (due to bad disks, cables, memory, or kernel bugs) going
260 unnoticed, ultimately resulting in data loss or corruption.
263 Add an ext3 journal to the filesystem. If the
265 option is not specified, the default journal parameters will be used to create
266 an appropriately sized journal (given the size of the filesystem)
267 stored within the filesystem. Note that you must be using a kernel
268 which has ext3 support in order to actually make use of the journal.
270 If this option is used to create a journal on a mounted filesystem, an
273 will be created in the top-level directory of the filesystem, as it is
274 the only safe way to create the journal inode while the filesystem is
275 mounted. While the ext3 journal is visible, it is not safe to
276 delete it, or modify it while the filesystem is mounted; for this
277 reason the file is marked immutable.
278 While checking unmounted filesystems,
280 will automatically move
282 files to the invisible, reserved journal inode. For all filesystems
283 except for the root filesystem, this should happen automatically and
284 naturally during the next reboot cycle. Since the root filesystem is
287 must be run from a rescue floppy in order to effect this transition.
289 On some distributions, such as Debian, if an initial ramdisk is used,
290 the initrd scripts will automatically convert an ext2 root filesystem
293 file specifies the ext3 filesystem for the root filesystem in order to
294 avoid requiring the use of a rescue floppy to add an ext3 journal to
297 .BR \-J " journal-options"
298 Override the default ext3 journal parameters. Journal options are comma
299 separated, and may take an argument using the equals ('=') sign.
300 The following journal options are supported:
303 .BI size= journal-size
304 Create a journal stored in the filesystem of size
306 megabytes. The size of the journal must be at least 1024 filesystem blocks
307 (i.e., 1MB if using 1k blocks, 4MB if using 4k blocks, etc.)
308 and may be no more than 102,400 filesystem blocks.
309 There must be enough free space in the filesystem to create a journal of
312 @JDEV@.BI device= external-journal
313 @JDEV@Attach the filesystem to the journal block device located on
314 @JDEV@.IR external-journal .
316 @JDEV@journal must have been already created using the command
318 @JDEV@.B mke2fs -O journal_dev
319 @JDEV@.I external-journal
322 @JDEV@.I external-journal
323 @JDEV@must be formatted with the same block
324 @JDEV@size as filesystems which will be using it.
325 @JDEV@In addition, while there is support for attaching
326 @JDEV@multiple filesystems to a single external journal,
327 @JDEV@the Linux kernel and
329 @JDEV@do not currently support shared external journals yet.
331 @JDEV@Instead of specifying a device name directly,
332 @JDEV@.I external-journal
333 @JDEV@can also be specified by either
334 @JDEV@.BI LABEL= label
337 @JDEV@to locate the external journal by either the volume label or UUID
338 @JDEV@stored in the ext2 superblock at the start of the journal. Use
339 @JDEV@.BR dumpe2fs (8)
340 @JDEV@to display a journal device's volume label and UUID. See also the
343 @JDEV@.BR tune2fs (8).
346 @JDEV@Only one of the
347 @JDEV@.BR size " or " device
348 @JDEV@options can be given for a filesystem.
351 List the contents of the filesystem superblock, including the current
352 values of the parameters that can be set via this program.
354 .BI \-L " volume-label"
355 Set the volume label of the filesystem.
356 Ext2 filesystem labels can be at most 16 characters long; if
358 is longer than 16 characters,
360 will truncate it and print a warning. The volume label can be used
366 (and possibly others) by specifying
367 .BI LABEL= volume_label
368 instead of a block special device name like
371 .BI \-m " reserved-blocks-percentage"
372 Set the percentage of the filesystem which may only be allocated
373 by privileged processes. Reserving some number of filesystem blocks
374 for use by privileged processes is done
375 to avoid filesystem fragmentation, and to allow system
378 to continue to function correctly after non-privileged processes are
379 prevented from writing to the filesystem. Normally, the default percentage
380 of reserved blocks is 5%.
382 .BI \-M " last-mounted-directory"
383 Set the last-mounted directory for the filesystem.
385 .BR \-o " [^]\fImount-option\fR[,...]"
386 Set or clear the indicated default mount options in the filesystem.
387 Default mount options can be overridden by mount options specified
390 or on the command line arguments to
392 Older kernels may not support this feature; in particular,
393 kernels which predate 2.4.20 will almost certainly ignore the
394 default mount options field in the superblock.
396 More than one mount option can be cleared or set by separating
397 features with commas. Mount options prefixed with a
398 caret character ('^') will be cleared in the filesystem's superblock;
399 mount options without a prefix character or prefixed with a plus
400 character ('+') will be added to the filesystem.
402 The following mount options can be set or cleared using
407 Enable debugging code for this filesystem.
410 Emulate BSD behaviour when creating new files: they will take the group-id
411 of the directory in which they were created. The standard System V behaviour
412 is the default, where newly created files take on the fsgid of the current
413 process, unless the directory has the setgid bit set, in which case it takes
414 the gid from the parent directory, and also gets the setgid bit set if it is
418 Enable user-specified extended attributes.
421 Enable Posix Access Control Lists.
424 Disables 32-bit UIDs and GIDs. This is for interoperability with
425 older kernels which only store and expect 16-bit values.
428 When the filesystem is mounted with journalling enabled, all data
429 (not just metadata) is committed into the journal prior to being written
430 into the main filesystem.
432 .B journal_data_ordered
433 When the filesystem is mounted with journalling enabled, all data is forced
434 directly out to the main file system prior to its metadata being committed
437 .B journal_data_writeback
438 When the filesystem is mounted with journalling enabled, data may be
439 written into the main filesystem after its metadata has been committed
440 to the journal. This may increase throughput, however, it may allow old
441 data to appear in files after a crash and journal recovery.
444 The file system will be mounted with barrier operations in the journal
445 disabled. (This option is currently only supported by the ext4 file
446 system driver in 2.6.35+ kernels.)
449 The file system will be mounted with the block_validity option enabled,
450 which causes extra checks to be performed after reading or writing from
451 the file system. This prevents corrupted metadata blocks from causing
452 file system damage by overwriting parts of the inode table or block
453 group descriptors. This comes at the cost of increased memory and CPU
454 overhead, so it is enabled only for debugging purposes. (This option is
455 currently only supported by the ext4 file system driver in 2.6.35+
459 The file system will be mouinted with the discard mount option. This will
460 cause the file system driver to attempt to use the trim/discard feature
461 of some storage devices (such as SSD's and thin-provisioned drives
462 available in some enterprise storage arrays) to inform the storage
463 device that blocks belonging to deleted files can be reused for other
464 purposes. (This option is currently only supported by the ext4 file
465 system driver in 2.6.35+ kernels.)
468 The file system will be mounted with the nodelalloc mount option. This
469 will disable the delayed allocation feature. (This option is currently
470 only supported by the ext4 file system driver in 2.6.35+ kernels.)
473 .BR \-O " [^]\fIfeature\fR[,...]"
474 Set or clear the indicated filesystem features (options) in the filesystem.
475 More than one filesystem feature can be cleared or set by separating
476 features with commas. Filesystem features prefixed with a
477 caret character ('^') will be cleared in the filesystem's superblock;
478 filesystem features without a prefix character or prefixed with a plus
479 character ('+') will be added to the filesystem.
481 The following filesystem features can be set or cleared using
486 Use hashed b-trees to speed up lookups in large directories.
489 Store file type information in directory entries.
492 Allow bitmaps and inode tables for a block group to be placed
493 anywhere on the storage media. \fBTune2fs\fR will not reorganize
494 the location of the inode tables and allocation bitmaps, as
496 will do when it creates a freshly formated file system with
501 Use a journal to ensure filesystem consistency even across unclean shutdowns.
502 Setting the filesystem feature is equivalent to using the
507 Filesystem can contain files that are greater than 2GB. (Modern kernels
508 set this feature automatically when a file > 2GB is created.)
511 Reserve space so the block group descriptor table may grow in the
514 only supports clearing this filesystem feature.
517 Limit the number of backup superblocks to save space on large filesystems.
520 Allow the kernel to initialize bitmaps and inode tables and keep a high
521 watermark for the unused inodes in a filesystem, to reduce
523 time. This first e2fsck run after enabling this feature will take the
524 full time, but subsequent e2fsck runs will take only a fraction of the
525 original time, depending on how full the file system is.
528 After setting or clearing
536 must be run on the filesystem to return the filesystem to a consistent state.
538 will print a message requesting that the system administrator run
540 if necessary. After setting the
544 can be run to convert existing directories to the hashed B-tree format.
545 Enabling certain filesystem features may prevent the filesystem from being
546 mounted by kernels which do not support those features. In particular, the
550 features are only supported by the ext4 filesystem.
552 .BI \-r " reserved-blocks-count"
553 Set the number of reserved filesystem blocks.
555 .BI \-T " time-last-checked"
556 Set the time the filesystem was last checked using
558 The time is interpreted using the current (local) timezone.
559 This can be useful in scripts which use a Logical Volume Manager to make
560 a consistent snapshot of a filesystem, and then check the filesystem
561 during off hours to make sure it hasn't been corrupted due to
562 hardware problems, etc. If the filesystem was clean, then this option can
563 be used to set the last checked time on the original filesystem. The format
566 is the international date format, with an optional time specifier, i.e.
567 YYYYMMDD[HH[MM[SS]]]. The keyword
569 is also accepted, in which case the last checked time will be set to the
573 Set the user who can use the reserved filesystem blocks.
575 can be a numerical uid or a user name. If a user name is given, it
576 is converted to a numerical uid before it is stored in the superblock.
579 Set the universally unique identifier (UUID) of the filesystem to
581 The format of the UUID is a series of hex digits separated by hyphens,
583 "c1b9d5a2-f162-11cf-9ece-0020afc76f16".
586 parameter may also be one of the following:
590 clear the filesystem UUID
593 generate a new randomly-generated UUID
596 generate a new time-based UUID
599 The UUID may be used by
604 (and possibly others) by specifying
606 instead of a block special device name like
611 for more information.
612 If the system does not have a good random number generator such as
617 will automatically use a time-based UUID instead of a randomly-generated UUID.
619 We haven't found any bugs yet. That doesn't mean there aren't any...
622 was written by Remy Card <Remy.Card@linux.org>. It is currently being
623 maintained by Theodore Ts'o <tytso@alum.mit.edu>.
625 uses the ext2fs library written by Theodore Ts'o <tytso@mit.edu>.
626 This manual page was written by Christian Kuhtz <chk@data-hh.Hanse.DE>.
627 Time-dependent checking was added by Uwe Ohse <uwe@tirka.gun.de>.
630 is part of the e2fsprogs package and is available from
631 http://e2fsprogs.sourceforge.net.