2 .\" Copyright (c) 2012, Martin Matuska <mm@FreeBSD.org>.
3 .\" All Rights Reserved.
5 .\" The contents of this file are subject to the terms of the
6 .\" Common Development and Distribution License (the "License").
7 .\" You may not use this file except in compliance with the License.
9 .\" You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
10 .\" or http://www.opensolaris.org/os/licensing.
11 .\" See the License for the specific language governing permissions
12 .\" and limitations under the License.
14 .\" When distributing Covered Code, include this CDDL HEADER in each
15 .\" file and include the License file at usr/src/OPENSOLARIS.LICENSE.
16 .\" If applicable, add the following below this CDDL HEADER, with the
17 .\" fields enclosed by brackets "[]" replaced with your own identifying
18 .\" information: Portions Copyright [yyyy] [name of copyright owner]
20 .\" Copyright (c) 2012 by Delphix. All rights reserved.
21 .\" Copyright (c) 2013 by Saso Kiselkov. All rights reserved.
22 .\" Copyright (c) 2013, Joyent, Inc. All rights reserved.
31 .Nd ZFS pool feature descriptions
33 ZFS pool on\-disk format versions are specified via "features" which replace
34 the old on\-disk format numbers (the last supported on\-disk format number is
36 To enable a feature on a pool use the
41 .Sy feature@feature_name
45 The pool format does not affect file system version compatibility or the ability
46 to send file systems between pools.
48 Since most features can be enabled independently of each other the on\-disk
49 format of the pool is specified by the set of all features marked as
51 on the pool. If the pool was created by another software version this set may
52 include unsupported features.
53 .Ss Identifying features
54 Every feature has a guid of the form
55 .Sy com.example:feature_name .
56 The reverse DNS name ensures that the feature's guid is unique across all ZFS
57 implementations. When unsupported features are encountered on a pool they will
58 be identified by their guids.
59 Refer to the documentation for the ZFS implementation that created the pool
60 for information about those features.
62 Each supported feature also has a short name.
63 By convention a feature's short name is the portion of its guid which follows
65 .Sy com.example:feature_name
66 would have the short name
68 however a feature's short name may differ across ZFS implementations if
69 following the convention would result in name conflicts.
71 Features can be in one of three states:
72 .Bl -tag -width "XXXXXXXX"
74 This feature's on\-disk format changes are in effect on the pool.
75 Support for this feature is required to import the pool in read\-write mode.
76 If this feature is not read-only compatible, support is also required to
77 import the pool in read\-only mode (see "Read\-only compatibility").
79 An administrator has marked this feature as enabled on the pool, but the
80 feature's on\-disk format changes have not been made yet.
81 The pool can still be imported by software that does not support this feature,
82 but changes may be made to the on\-disk format at any time which will move
86 Some features may support returning to the
90 See feature\-specific documentation for details.
92 This feature's on\-disk format changes have not been made and will not be made
93 unless an administrator moves the feature to the
96 Features cannot be disabled once they have been enabled.
99 The state of supported features is exposed through pool properties of the form
100 .Sy feature@short_name .
101 .Ss Read\-only compatibility
102 Some features may make on\-disk format changes that do not interfere with other
103 software's ability to read from the pool.
104 These features are referred to as "read\-only compatible".
105 If all unsupported features on a pool are read\-only compatible, the pool can
106 be imported in read\-only mode by setting the
108 property during import (see
110 for details on importing pools).
111 .Ss Unsupported features
112 For each unsupported feature enabled on an imported pool a pool property
114 .Sy unsupported@feature_guid
115 will indicate why the import was allowed despite the unsupported feature.
116 Possible values for this property are:
117 .Bl -tag -width "XXXXXXXX"
119 The feature is in the
121 state and therefore the pool's on\-disk format is still compatible with
122 software that does not support this feature.
124 The feature is read\-only compatible and the pool has been imported in
127 .Ss Feature dependencies
128 Some features depend on other features being enabled in order to function
130 Enabling a feature will automatically enable any features it depends on.
132 The following features are supported on this system:
133 .Bl -tag -width "XXXXXXXX"
135 .Bl -column "READ\-ONLY COMPATIBLE" "com.delphix:async_destroy"
136 .It GUID Ta com.delphix:async_destroy
137 .It READ\-ONLY COMPATIBLE Ta yes
138 .It DEPENDENCIES Ta none
141 Destroying a file system requires traversing all of its data in order to
142 return its used space to the pool.
145 the file system is not fully removed until all space has been reclaimed.
146 If the destroy operation is interrupted by a reboot or power outage the next
147 attempt to open the pool will need to complete the destroy operation
152 is enabled the file system's data will be reclaimed by a background process,
153 allowing the destroy operation to complete without traversing the entire file
155 The background process is able to resume interrupted destroys after the pool
156 has been opened, eliminating the need to finish interrupted destroys as part
157 of the open operation.
158 The amount of space remaining to be reclaimed by the background process is
159 available through the
169 .Bl -column "READ\-ONLY COMPATIBLE" "com.delphix:empty_bpobj"
170 .It GUID Ta com.delphix:empty_bpobj
171 .It READ\-ONLY COMPATIBLE Ta yes
172 .It DEPENDENCIES Ta none
175 This feature increases the performance of creating and using a large number
176 of snapshots of a single filesystem or volume, and also reduces the disk
179 When there are many snapshots, each snapshot uses many Block Pointer Objects
181 to track blocks associated with that snapshot.
182 However, in common use cases, most of these bpobj's are empty.
183 This feature allows us to create each bpobj on-demand, thus eliminating the
188 while there are any filesystems, volumes, or snapshots which were created
189 after enabling this feature.
190 .It Sy filesystem_limits
191 .Bl -column "READ\-ONLY COMPATIBLE" "com.joyent:filesystem_limits"
192 .It GUID Ta com.joyent:filesystem_limits
193 .It READ\-ONLY COMPATIBLE Ta yes
194 .It DEPENDENCIES Ta extensible_dataset
197 This feature enables filesystem and snapshot limits.
198 These limits can be used
199 to control how many filesystems and/or snapshots can be created at the point in
200 the tree on which the limits are set.
204 once either of the limit properties has been
206 Once activated the feature is never deactivated.
208 .Bl -column "READ\-ONLY COMPATIBLE" "org.illumos:lz4_compress"
209 .It GUID Ta org.illumos:lz4_compress
210 .It READ\-ONLY COMPATIBLE Ta no
211 .It DEPENDENCIES Ta none
215 is a high-performance real-time compression algorithm that
216 features significantly faster compression and decompression as well as a
217 higher compression ratio than the older
222 compression is approximately 50% faster on
223 compressible data and 200% faster on incompressible data than
225 It is also approximately 80% faster on decompression, while
226 giving approximately 10% better compression ratio.
233 administrator can turn on
235 compression on any dataset on the
239 Also, all newly written metadata
240 will be compressed with
243 Since this feature is not read-only compatible, this
244 operation will render the pool unimportable on systems without support
250 -compressed root pools is supported.
254 as soon as it is enabled and will
255 never return to being
257 .It Sy multi_vdev_crash_dump
258 .Bl -column "READ\-ONLY COMPATIBLE" "com.joyent:multi_vdev_crash_dump"
259 .It GUID Ta com.joyent:multi_vdev_crash_dump
260 .It READ\-ONLY COMPATIBLE Ta no
261 .It DEPENDENCIES Ta none
264 This feature allows a dump device to be configured with a pool comprised
266 Those vdevs may be arranged in any mirrored or raidz
268 .\" TODO: this is not yet supported on FreeBSD.
271 .\" .Sy multi_vdev_crash_dump
272 .\" feature is set to
274 .\" the administrator can use the
276 .\" command to configure a
277 .\" dump device on a pool comprised of multiple vdevs.
278 .It Sy spacemap_histogram
279 .Bl -column "READ\-ONLY COMPATIBLE" "com.delphix:spacemap_histogram"
280 .It GUID Ta com.delphix:spacemap_histogram
281 .It READ\-ONLY COMPATIBLE Ta yes
282 .It DEPENDENCIES Ta none
285 This features allows ZFS to maintain more information about how free space
286 is organized within the pool. If this feature is
291 when a new space map object is created or
292 an existing space map is upgraded to the new format.
295 it will remain in that state until the pool is destroyed.
296 .It Sy extensible_dataset
297 .Bl -column "READ\-ONLY COMPATIBLE" "com.delphix:extensible_dataset"
298 .It GUID Ta com.delphix:extensible_dataset
299 .It READ\-ONLY COMPATIBLE Ta no
300 .It DEPENDENCIES Ta none
303 This feature allows more flexible use of internal ZFS data structures,
304 and exists for other features to depend on.
308 when the first dependent feature uses it,
309 and will be returned to the
311 state when all datasets that use
312 this feature are destroyed.
314 .Bl -column "READ\-ONLY COMPATIBLE" "com.delphix:bookmarks"
315 .It GUID Ta com.delphix:bookmarks
316 .It READ\-ONLY COMPATIBLE Ta yes
317 .It DEPENDENCIES Ta extensible_dataset
320 This feature enables use of the
327 while any bookmarks exist in the pool.
328 All bookmarks in the pool can be listed by running
331 .Fl t No bookmark Fl r Ar poolname .
333 .Bl -column "READ\-ONLY COMPATIBLE" "com.delphix:enabled_txg"
334 .It GUID Ta com.delphix:enabled_txg
335 .It READ\-ONLY COMPATIBLE Ta yes
336 .It DEPENDENCIES Ta none
339 Once this feature is enabled ZFS records the transaction group number
340 in which new features are enabled. This has no user-visible impact,
341 but other features may depend on this feature.
345 as soon as it is enabled and will
346 never return to being
349 .Bl -column "READ\-ONLY COMPATIBLE" "com.delphix:hole_birth"
350 .It GUID Ta com.delphix:hole_birth
351 .It READ\-ONLY COMPATIBLE Ta no
352 .It DEPENDENCIES Ta enabled_txg
355 This feature improves performance of incremental sends
357 and receives for objects with many holes.
358 The most common case of
359 hole-filled objects is zvols.
361 An incremental send stream from snapshot
365 contains information about every block that changed between
369 Blocks which did not change between those snapshots can be
370 identified and omitted from the stream using a piece of metadata called
371 the 'block birth time', but birth times are not recorded for holes
372 .Pq blocks filled only with zeroes .
373 Since holes created after
376 distinguished from holes created before
378 information about every
379 hole in the entire filesystem or zvol is included in the send stream.
381 For workloads where holes are rare this is not a problem.
383 incrementally replicating filesystems or zvols with many holes
384 .Pq for example a zvol formatted with another filesystem
386 be spent sending and receiving unnecessary information about holes that
387 already exist on the receiving side.
391 feature has been enabled the block birth times
392 of all new holes will be recorded.
393 Incremental sends between snapshots
394 created after this feature is enabled will use this new metadata to avoid
395 sending information about holes that already exist on the receiving side.
399 as soon as it is enabled and will
400 never return to being
403 .Bl -column "READ\-ONLY COMPATIBLE" "com.delphix:embedded_data"
404 .It GUID Ta com.delphix:embedded_data
405 .It READ\-ONLY COMPATIBLE Ta no
406 .It DEPENDENCIES Ta none
409 This feature improves the performance and compression ratio of
410 highly-compressible blocks.
411 Blocks whose contents can compress to 112 bytes
412 or smaller can take advantage of this feature.
414 When this feature is enabled, the contents of highly-compressible blocks are
415 stored in the block "pointer" itself
416 .Po a misnomer in this case, as it contains
417 the compressed data, rather than a pointer to its location on disk
420 the space of the block
421 .Pq one sector, typically 512 bytes or 4KB
423 and no additional i/o is needed to read and write the data block.
427 as soon as it is enabled and will
428 never return to being
431 .Bl -column "READ\-ONLY COMPATIBLE" "org.open-zfs:large_block"
432 .It GUID Ta org.open-zfs:large_block
433 .It READ\-ONLY COMPATIBLE Ta no
434 .It DEPENDENCIES Ta extensible_dataset
439 feature allows the record size on a dataset to be
440 set larger than 128KB.
446 property has been set larger than 128KB, and will return to being
448 once all filesystems that have ever had their recordsize larger than 128KB
451 Please note that booting from datasets that have recordsize greater than
461 This manual page is a
463 reimplementation of the
466 .Em zpool-features(5) ,
467 modified and customized for
469 and licensed under the Common Development and Distribution License
474 implementation of this manual page was initially written by
475 .An Martin Matuska Aq mm@FreeBSD.org .