1mkfs.ocfs2(8) OCFS2 Manual Pages mkfs.ocfs2(8)
2
6 mkfs.ocfs2 - Creates an OCFS2 file system.
7
9 mkfs.ocfs2 [-b block-size] [-C cluster-size] [-L volume-label] [-M
10 mount-type] [-N number-of-nodes] [-J journal-options] [--fs-fea‐
11 tures=[no]sparse...] [--fs-feature-level=feature-level] [-T filesys‐
12 tem-type] [--cluster-stack=stackname] [--cluster-name=clustername]
13 [--global-heartbeat] [--discard | --nodiscard] [-FqvV] device [blocks-
14 count]
15
17 mkfs.ocfs2 is used to create an OCFS2 file system on a device, usually
18 a partition on a shared disk. In order to prevent data loss, mkfs.ocfs2
19 will not format an existing OCFS2 volume if it detects that it is
20 mounted on another node in the cluster. This tool requires the cluster
21 service to be online.
22
25 -b, --block-size block-size
26 Valid block size values are 512, 1K, 2K and 4K bytes per block.
27 If omitted, a value will be heuristically determined based on
28 the expected usage of the file system (see the -T option). A
29 block size of 512 bytes is never recommended. Choose 1K, 2K or
30 4K.
31 -C, --cluster-size cluster-size
34 Valid cluster size values are 4K, 8K, 16K, 32K, 64K, 128K, 256K,
35 512K and 1M. If omitted, a value will be heuristically deter‐
36 mined based on the expected usage of the file system (see the -T
37 option). For volumes expected to store large files, like data‐
38 base files, while a cluster size of 128K or more is recommended,
39 one can opt for a smaller size as long as that value is not
40 smaller than the database block size. For others, use 4K.
41 -F, --force
44 For existing OCFS2 volumes, mkfs.ocfs2 ensures the volume is not
45 mounted on any node in the cluster before formatting. For that
46 to work, mkfs.ocfs2 expects the cluster service to be online.
47 Specify this option to disable this check.
48 -J, --journal-options options
51 Create the journal using options specified on the command-line.
52 Journal options are comma separated, and may take an argument
53 using the equals ('=') sign. The following options are sup‐
54 ported:
55 size=journal-size
58 Create a journal of size journal-size. Minimum size is
59 4M. If omitted, a value is heuristically determined
60 based upon the file system size.
61 block32
64 Use a standard 32bit journal. The journal will be able
65 to access up to 2^32-1 blocks. This is the default. It
66 has been the journal format for OCFS2 volumes since the
67 beginning. The journal is compatible with all versions
68 of OCFS2. Prepending no is equivalent to the block64
69 journal option.
70 block64
73 Use a 64bit journal. The journal will be able to access
74 up to 2^64-1 blocks. This allows large filesystems that
75 can extend to the theoretical limits of OCFS2. It re‐
76 quires a new-enough filesystem driver that uses the new
77 journalled block device, JBD2. Prepending no is equiva‐
78 lent to the block32 journal option.
79 -L, --label volume-label
82 Set the volume label for the file system. This is useful for
83 mounting-by-label. Limit the label to under 64 bytes.
84 -M, --mount mount-type
87 Valid types are local and cluster. Local mount allows users to
88 mount the volume without the cluster overhead and works only
89 with OCFS2 bundled with Linux kernels 2.6.20 or later. Defaults
90 to cluster.
91 -N, --node-slots number-of-node-slots
94 Valid number ranges from 1 to 255. This number specifies the
95 maximum number of nodes that can concurrently mount the parti‐
96 tion. If omitted, the number depends on volume size, for volume
97 size < 2G, it's 2, for 2G <= size < 10G, it's 4, for 10G <= size
98 < 1T, it's 8, for other size, it's 16. The number of slots can
99 be later tuned up or down using tunefs.ocfs2.
100 -T filesystem-type
103 Specify how the filesystem is going to be used, so that
104 mkfs.ocfs2 can chose optimal filesystem parameters for that use.
105 The supported filesystem types are:
106 mail Appropriate for file systems that will host lots of small
109 files.
110 datafiles
113 Appropriate for file systems that will host a relatively
114 small number of very large files.
115 vmstore
118 Appropriate for file systems that will host Virtual ma‐
119 chine images.
120 --fs-features=[no]sparse...
123 Turn specific file system features on or off. A comma separated
124 list of feature flags can be provided, and mkfs.ocfs2 will try
125 to create the file system with those features set according to
126 the list. To turn a feature on, include it in the list. To turn
127 a feature off, prepend no to the name. Choices here will over‐
128 ride individual features set via the --fs-feature-level option.
129 Refer to the section titled feature compatibility before select‐
130 ing specific features. The following flags are supported:
131 backup-super
134 mkfs.ocfs2, by default, makes up to 6 backup copies of
135 the super block at offsets 1G, 4G, 16G, 64G, 256G and 1T
136 depending on the size of the volume. This can be useful
137 in disaster recovery. This feature is fully compatible
138 with all versions of the file system and generally should
139 not be disabled.
140 local Create the file system as a local mount, so that it can
143 be mounted without a cluster stack.
144 sparse Enable support for sparse files. With this, OCFS2 can
147 avoid allocating (and zeroing) data to fill holes. Turn
148 this feature on if you can, otherwise extends and some
149 writes might be less performant.
150 unwritten
153 Enable unwritten extents support. With this turned on, an
154 application can request that a range of clusters be pre-
155 allocated within a file. OCFS2 will mark those extents
156 with a special flag so that expensive data zeroing
157 doesn't have to be performed. Reads and writes to a pre-
158 allocated region act as reads and writes to a hole, ex‐
159 cept a write will not fail due to lack of data alloca‐
160 tion. This feature requires sparse file support to be
161 turned on.
162 inline-data
165 Enable inline-data support. If this feature is turned on,
166 OCFS2 will store small files and directories inside the
167 inode block. Data is transparently moved out to an extent
168 when it no longer fits inside the inode block. In some
169 cases, this can also make a positive impact on cold-cache
170 directory and file operations.
171 extended-slotmap
174 The slot-map is a hidden file on an OCFS2 fs which is
175 used to map mounted nodes to system file resources. The
176 extended slot map allows a larger range of possible node
177 numbers, which is useful for userspace cluster stacks. If
178 required, this feature is automatically turned on by
179 mkfs.ocfs2.
180 metaecc
183 Enables metadata checksums. With this enabled, the file
184 system computes and stores the checksums in all metadata
185 blocks. It also computes and stores an error correction
186 code capable of fixing single bit errors.
187 refcount
190 Enables creation of reference counted trees. With this
191 enabled, the file system allows users to create inode-
192 based snapshots and clones known as reflinks.
193 xattr Enable extended attributes support. With this enabled,
196 users can attach name:value pairs to objects within the
197 file system. In OCFS2, the names can be up to 255 bytes
198 in length, terminated by the first NUL byte. While it is
199 not required, printable names (ASCII) are recommended.
200 The values can be up to 64KB of arbitrary binary data.
201 Attributes can be attached to all types of inodes: regu‐
202 lar files, directories, symbolic links, device nodes,
203 etc. This feature is required for users wanting to use
204 extended security facilities like POSIX ACLs or SELinux.
205 usrquota
208 Enable user quota support. With this feature enabled,
209 filesystem will track amount of space and number of in‐
210 odes (files, directories, symbolic links) each user owns.
211 It is then possible to limit the maximum amount of space
212 or inodes user can have. See a documentation of quota-
213 tools package for more details.
214 grpquota
217 Enable group quota support. With this feature enabled,
218 filesystem will track amount of space and number of in‐
219 odes (files, directories, symbolic links) each group
220 owns. It is then possible to limit the maximum amount of
221 space or inodes user can have. See a documentation of
222 quota-tools package for more details.
223 indexed-dirs
226 Enable directory indexing support. With this feature en‐
227 abled, the file system creates indexed tree for non-in‐
228 line directory entries. For large scale directories, di‐
229 rectory entry lookup performance from the indexed tree is
230 faster then from the legacy directory blocks.
231 discontig-bg
234 Enables discontiguous block groups. With this feature en‐
235 abled, the file system is able to grow the inode and the
236 extent allocators even when there is no contiguous free
237 chunk available. It allows the file system to grow the
238 allocators in smaller (discontiguous) chunks.
239 clusterinfo
242 Enables storing the cluster stack information in the su‐
243 perblock. This feature is needed to support userspace
244 cluster stacks and the global heartbeat mode in the o2cb
245 cluster stack. If needed, this feature is automatically
246 turned on by mkfs.ocfs2.
247 --fs-feature-level=feature-level
250 Choose from a set of pre-determined file-system features. This
251 option is designed to allow users to conveniently choose a set
252 of file system features which fits their needs. There is no
253 downside to trying a set of features which your module might not
254 support - if it won't mount the new file system simply reformat
255 at a lower level. Feature levels can be fine-tuned via the
256 --fs-features option. Currently, there are 3 types of feature
257 levels:
258 max-compat
261 Chooses fewer features but ensures that the file system
262 can be mounted from older versions of the OCFS2 module.
263 default
266 The default feature set tries to strike a balance between
267 providing new features and maintaining compatibility with
268 relatively recent versions of OCFS2. It currently enables
269 sparse, unwritten, inline-data, xattr, indexed-dirs, dis‐
270 contig-bg, refcount, extended-slotmap and clusterinfo.
271 max-features
274 Choose the maximum amount of features available. This
275 will typically provide the best performance from OCFS2 at
276 the expense of creating a file system that is only com‐
277 patible with very recent versions of the OCFS2 kernel
278 module.
279 --cluster-stack
282 Specify the cluster stack. This option is normally not required
283 as mkfs.ocfs2 chooses the currently active cluster stack. It is
284 required only if the cluster stack is not online and the user
285 wishes to use a stack other than the default, o2cb. Other sup‐
286 ported cluster stacks are pcmk (Pacemaker) and cman (rgmanager).
287 Once set, OCFS2 will only allow mounting the volume if the ac‐
288 tive cluster stack and cluster name matches the one specified
289 on-disk.
290 --cluster-name
293 Specify the name of the cluster. This option is mandatory if the
294 user has specified a cluster-stack. This name is restricted to a
295 max of 16 characters. Additionally, the o2cb cluster stack al‐
296 lows only alpha-numeric characters.
297 --global-heartbeat
300 Enable the global heartbeat mode of the o2cb cluster stack. This
301 option is not required if the o2cb cluster stack with global
302 heartbeat is online as mkfs.ocfs2 will detect the active stack.
303 However, if the cluster stack is not up, then this option is re‐
304 quired alongwith cluster-stack and cluster-name. For more, re‐
305 fer to o2cb(7).
306 --discard
309 Attempt to discard blocks at mkfs time (discarding blocks ini‐
310 tially is useful on solid state devices and sparse / thin-provi‐
311 sioned storage). When the device advertises that discard also
312 zeroes data (any subsequent read after the discard and before
313 write returns zero), then mark all not-yet-zeroed blocks as ze‐
314 roed. This significantly speeds up filesystem initialization.
315 This is set as default.
316 --nodiscard
319 Do not attempt to discard blocks at mkfs time.
320 --no-backup-super
323 This option is deprecated, please use --fs-features=nobackup-su‐
324 per instead.
325 -n, --dry-run
328 Display the heuristically determined values without overwriting
329 the existing file system.
330 -q, --quiet
333 Quiet mode.
334 -U uuid
337 Specify a custom UUID in the plain
338 (2A4D1C581FAA42A1A41D26EFC90C1315) or traditional
339 (2a4d1c58-1faa-42a1-a41d-26efc90c1315) format. This option in
340 not recommended because the file system uses the UUID to
341 uniquely identify a file system. If more than one file system
342 were to have the same UUID, one is very likely to encounter er‐
343 ratic behavior, if not, outright file system corruption.
344 -v, --verbose
347 Verbose mode.
348 -V, --version
351 Print version and exit.
352 blocks-count
355 Usually mkfs.ocfs2 automatically determines the size of the
356 given device and creates a file system that uses all of the
357 available space on the device. This optional argument specifies
358 that the file system should only consume the given number of
359 file system blocks (see -b) on the device.
360
363 This section lists the file system features that have been added to the
364 OCFS2 file system and the version that it first appeared in. The table
365 below lists the versions of the mainline Linux kernel and ocfs2-tools.
366 Users should use this information to enable only those features that
367 are available in the file system that they are using. Before enabling
368 new features, users are advised to review to the section titled feature
369 values.
370 ┌─────────────────┬────────────────┬─────────────────┐
373 │Feature │ Kernel Version │ Tools Version │
374 ├─────────────────┼────────────────┼─────────────────┤
375 │local │ Linux 2.6.20 │ ocfs2-tools 1.2 │
376 ├─────────────────┼────────────────┼─────────────────┤
377 │sparse │ Linux 2.6.22 │ ocfs2-tools 1.4 │
378 ├─────────────────┼────────────────┼─────────────────┤
379 │unwritten │ Linux 2.6.23 │ ocfs2-tools 1.4 │
380 ├─────────────────┼────────────────┼─────────────────┤
381 │inline-data │ Linux 2.6.24 │ ocfs2-tools 1.4 │
382 ├─────────────────┼────────────────┼─────────────────┤
383 │extended-slotmap │ Linux 2.6.27 │ ocfs2-tools 1.6 │
384 ├─────────────────┼────────────────┼─────────────────┤
385 │metaecc │ Linux 2.6.29 │ ocfs2-tools 1.6 │
386 ├─────────────────┼────────────────┼─────────────────┤
387 │grpquota │ Linux 2.6.29 │ ocfs2-tools 1.6 │
388 ├─────────────────┼────────────────┼─────────────────┤
389 │usrquota │ Linux 2.6.29 │ ocfs2-tools 1.6 │
390 ├─────────────────┼────────────────┼─────────────────┤
391 │xattr │ Linux 2.6.29 │ ocfs2-tools 1.6 │
392 ├─────────────────┼────────────────┼─────────────────┤
393 │indexed-dirs │ Linux 2.6.30 │ ocfs2-tools 1.6 │
394 ├─────────────────┼────────────────┼─────────────────┤
395 │refcount │ Linux 2.6.32 │ ocfs2-tools 1.6 │
396 ├─────────────────┼────────────────┼─────────────────┤
397 │discontig-bg │ Linux 2.6.35 │ ocfs2-tools 1.6 │
398 ├─────────────────┼────────────────┼─────────────────┤
399 │clusterinfo │ Linux 2.6.37 │ ocfs2-tools 1.8 │
400 └─────────────────┴────────────────┴─────────────────┘
401 Users can query the features enabled in the file system as follows:
403 # tunefs.ocfs2 -Q "Label: %V\nFeatures: %H %O\n" /dev/sdg1
405 Label: apache_files_10
406 Features: sparse inline-data unwritten
407
410 This section lists the hex values that are associated with the file
411 system features. This information is useful when debugging mount fail‐
412 ures that are due to feature incompatibility. When a user attempts to
413 mount an OCFS2 volume that has features enabled that are not supported
414 by the running file system software, it will fail with an error like:
415 ERROR: couldn't mount because of unsupported optional features (200).
417 By referring to the table below, it becomes apparent that the user at‐
419 tempted to mount a volume with the xattr (extended attributes) feature
420 enabled with a version of the file system software that did not support
421 it. At this stage, the user has the option of either upgrading the file
422 system software, or, disabling that on-disk feature using tunefs.ocfs2.
423 Some features allow the file system to be mounted with an older version
425 of the software provided the mount is read-only. If a user attempts to
426 mount such a volume in a read-write mode, it will fail with an error
427 like:
428 ERROR: couldn't mount RDWR because of unsupported optional features (1).
430 This error indicates that the volume had the unwritten RO compat fea‐
432 ture enabled. This volume can be mounted by an older file system soft‐
433 ware only in the read-only mode. In this case, the user has the option
434 of either mounting the volume with the ro mount option, or, disabling
435 that on-disk feature using tunefs.ocfs2.
436 ┌─────────────────┬───────────┬───────────┐
439 │Feature │ Category │ Hex value │
440 ├─────────────────┼───────────┼───────────┤
441 │local │ Incompat │ 8 │
442 ├─────────────────┼───────────┼───────────┤
443 │sparse │ Incompat │ 10 │
444 ├─────────────────┼───────────┼───────────┤
445 │inline-data │ Incompat │ 40 │
446 ├─────────────────┼───────────┼───────────┤
447 │extended-slotmap │ Incompat │ 100 │
448 ├─────────────────┼───────────┼───────────┤
449 │xattr │ Incompat │ 200 │
450 ├─────────────────┼───────────┼───────────┤
451 │indexed-dirs │ Incompat │ 400 │
452 ├─────────────────┼───────────┼───────────┤
453 │metaecc │ Incompat │ 800 │
454 ├─────────────────┼───────────┼───────────┤
455 │refcount │ Incompat │ 1000 │
456 ├─────────────────┼───────────┼───────────┤
457 │discontig-bg │ Incompat │ 2000 │
458 ├─────────────────┼───────────┼───────────┤
459 │clusterinfo │ Incompat │ 4000 │
460 ├─────────────────┼───────────┼───────────┤
461 │unwritten │ RO Compat │ 1 │
462 ├─────────────────┼───────────┼───────────┤
463 │usrquota │ RO Compat │ 2 │
464 ├─────────────────┼───────────┼───────────┤
465 │grpquota │ RO Compat │ 4 │
466 └─────────────────┴───────────┴───────────┘
467
470 debugfs.ocfs2(8) fsck.ocfs2(8) mount.ocfs2(8) mounted.ocfs2(8) o2cb(7)
471 o2cluster(8) o2image(8) o2info(1) tunefs.ocfs2(8)
472
475 Oracle Corporation
476
479 Copyright © 2004, 2012 Oracle. All rights reserved.
480Version 1.8.7 January 2012 mkfs.ocfs2(8)