man/man5/zpool-features.5

   1 '\" te
   2 .\" Copyright (c) 2012, 2015 by Delphix. All rights reserved.
   3 .\" Copyright (c) 2013 by Saso Kiselkov. All rights reserved.
   4 .\" Copyright (c) 2014, Joyent, Inc. All rights reserved.
   5 .\" The contents of this file are subject to the terms of the Common Development
   6 .\" and Distribution License (the "License").  You may not use this file except
   7 .\" in compliance with the License. You can obtain a copy of the license at
   8 .\" usr/src/OPENSOLARIS.LICENSE or http://www.opensolaris.org/os/licensing.
   9 .\"
  10 .\" See the License for the specific language governing permissions and
  11 .\" limitations under the License. When distributing Covered Code, include this
  12 .\" CDDL HEADER in each file and include the License file at
  13 .\" usr/src/OPENSOLARIS.LICENSE.  If applicable, add the following below this
  14 .\" CDDL HEADER, with the fields enclosed by brackets "[]" replaced with your
  15 .\" own identifying information:
  16 .\" Portions Copyright [yyyy] [name of copyright owner]
  17 .TH ZPOOL-FEATURES 5 "Aug 27, 2013"
  18 .SH NAME
  19 zpool\-features \- ZFS pool feature descriptions
  20 .SH DESCRIPTION
  21 .sp
  22 .LP
  23 ZFS pool on\-disk format versions are specified via "features" which replace
  24 the old on\-disk format numbers (the last supported on\-disk format number is
  25 28). To enable a feature on a pool use the \fBupgrade\fR subcommand of the
  26 \fBzpool\fR(8) command, or set the \fBfeature@\fR\fIfeature_name\fR property
  27 to \fBenabled\fR.
  28 .sp
  29 .LP
  30 The pool format does not affect file system version compatibility or the ability
  31 to send file systems between pools.
  32 .sp
  33 .LP
  34 Since most features can be enabled independently of each other the on\-disk
  35 format of the pool is specified by the set of all features marked as
  36 \fBactive\fR on the pool. If the pool was created by another software version
  37 this set may include unsupported features.
  38 .SS "Identifying features"
  39 .sp
  40 .LP
  41 Every feature has a guid of the form \fIcom.example:feature_name\fR. The reverse
  42 DNS name ensures that the feature's guid is unique across all ZFS
  43 implementations. When unsupported features are encountered on a pool they will
  44 be identified by their guids. Refer to the documentation for the ZFS
  45 implementation that created the pool for information about those features.
  46 .sp
  47 .LP
  48 Each supported feature also has a short name. By convention a feature's short
  49 name is the portion of its guid which follows the ':' (e.g.
  50 \fIcom.example:feature_name\fR would have the short name \fIfeature_name\fR),
  51 however a feature's short name may differ across ZFS implementations if
  52 following the convention would result in name conflicts.
  53 .SS "Feature states"
  54 .sp
  55 .LP
  56 Features can be in one of three states:
  57 .sp
  58 .ne 2
  59 .na
  60 \fB\fBactive\fR\fR
  61 .ad
  62 .RS 12n
  63 This feature's on\-disk format changes are in effect on the pool. Support for
  64 this feature is required to import the pool in read\-write mode. If this
  65 feature is not read-only compatible, support is also required to import the pool
  66 in read\-only mode (see "Read\-only compatibility").
  67 .RE
  68
  69 .sp
  70 .ne 2
  71 .na
  72 \fB\fBenabled\fR\fR
  73 .ad
  74 .RS 12n
  75 An administrator has marked this feature as enabled on the pool, but the
  76 feature's on\-disk format changes have not been made yet. The pool can still be
  77 imported by software that does not support this feature, but changes may be made
  78 to the on\-disk format at any time which will move the feature to the
  79 \fBactive\fR state. Some features may support returning to the \fBenabled\fR
  80 state after becoming \fBactive\fR. See feature\-specific documentation for
  81 details.
  82 .RE
  83
  84 .sp
  85 .ne 2
  86 .na
  87 \fBdisabled\fR
  88 .ad
  89 .RS 12n
  90 This feature's on\-disk format changes have not been made and will not be made
  91 unless an administrator moves the feature to the \fBenabled\fR state. Features
  92 cannot be disabled once they have been enabled.
  93 .RE
  94
  95 .sp
  96 .LP
  97 The state of supported features is exposed through pool properties of the form
  98 \fIfeature@short_name\fR.
  99 .SS "Read\-only compatibility"
 100 .sp
 101 .LP
 102 Some features may make on\-disk format changes that do not interfere with other
 103 software's ability to read from the pool. These features are referred to as
 104 "read\-only compatible". If all unsupported features on a pool are read\-only
 105 compatible, the pool can be imported in read\-only mode by setting the
 106 \fBreadonly\fR property during import (see \fBzpool\fR(8) for details on
 107 importing pools).
 108 .SS "Unsupported features"
 109 .sp
 110 .LP
 111 For each unsupported feature enabled on an imported pool a pool property
 112 named \fIunsupported@feature_guid\fR will indicate why the import was allowed
 113 despite the unsupported feature. Possible values for this property are:
 114
 115 .sp
 116 .ne 2
 117 .na
 118 \fB\fBinactive\fR\fR
 119 .ad
 120 .RS 12n
 121 The feature is in the \fBenabled\fR state and therefore the pool's on\-disk
 122 format is still compatible with software that does not support this feature.
 123 .RE
 124
 125 .sp
 126 .ne 2
 127 .na
 128 \fB\fBreadonly\fR\fR
 129 .ad
 130 .RS 12n
 131 The feature is read\-only compatible and the pool has been imported in
 132 read\-only mode.
 133 .RE
 134
 135 .SS "Feature dependencies"
 136 .sp
 137 .LP
 138 Some features depend on other features being enabled in order to function
 139 properly. Enabling a feature will automatically enable any features it
 140 depends on.
 141 .SH FEATURES
 142 .sp
 143 .LP
 144 The following features are supported on this system:
 145 .sp
 146 .ne 2
 147 .na
 148 \fB\fBasync_destroy\fR\fR
 149 .ad
 150 .RS 4n
 151 .TS
 152 l l .
 153 GUID    com.delphix:async_destroy
 154 READ\-ONLY COMPATIBLE   yes
 155 DEPENDENCIES    none
 156 .TE
 157
 158 Destroying a file system requires traversing all of its data in order to
 159 return its used space to the pool. Without \fBasync_destroy\fR the file system
 160 is not fully removed until all space has been reclaimed. If the destroy
 161 operation is interrupted by a reboot or power outage the next attempt to open
 162 the pool will need to complete the destroy operation synchronously.
 163
 164 When \fBasync_destroy\fR is enabled the file system's data will be reclaimed
 165 by a background process, allowing the destroy operation to complete without
 166 traversing the entire file system. The background process is able to resume
 167 interrupted destroys after the pool has been opened, eliminating the need
 168 to finish interrupted destroys as part of the open operation. The amount
 169 of space remaining to be reclaimed by the background process is available
 170 through the \fBfreeing\fR property.
 171
 172 This feature is only \fBactive\fR while \fBfreeing\fR is non\-zero.
 173 .RE
 174
 175 .sp
 176 .ne 2
 177 .na
 178 \fB\fBempty_bpobj\fR\fR
 179 .ad
 180 .RS 4n
 181 .TS
 182 l l .
 183 GUID    com.delphix:empty_bpobj
 184 READ\-ONLY COMPATIBLE   yes
 185 DEPENDENCIES    none
 186 .TE
 187
 188 This feature increases the performance of creating and using a large
 189 number of snapshots of a single filesystem or volume, and also reduces
 190 the disk space required.
 191
 192 When there are many snapshots, each snapshot uses many Block Pointer
 193 Objects (bpobj's) to track blocks associated with that snapshot.
 194 However, in common use cases, most of these bpobj's are empty.  This
 195 feature allows us to create each bpobj on-demand, thus eliminating the
 196 empty bpobjs.
 197
 198 This feature is \fBactive\fR while there are any filesystems, volumes,
 199 or snapshots which were created after enabling this feature.
 200 .RE
 201
 202 .sp
 203 .ne 2
 204 .na
 205 \fB\fBfilesystem_limits\fR\fR
 206 .ad
 207 .RS 4n
 208 .TS
 209 l l .
 210 GUID    com.joyent:filesystem_limits
 211 READ\-ONLY COMPATIBLE   yes
 212 DEPENDENCIES    extensible_dataset
 213 .TE
 214
 215 This feature enables filesystem and snapshot limits. These limits can be used
 216 to control how many filesystems and/or snapshots can be created at the point in
 217 the tree on which the limits are set.
 218
 219 This feature is \fBactive\fR once either of the limit properties has been
 220 set on a dataset. Once activated the feature is never deactivated.
 221 .RE
 222
 223 .sp
 224 .ne 2
 225 .na
 226 \fB\fBlz4_compress\fR\fR
 227 .ad
 228 .RS 4n
 229 .TS
 230 l l .
 231 GUID    org.illumos:lz4_compress
 232 READ\-ONLY COMPATIBLE   no
 233 DEPENDENCIES    none
 234 .TE
 235
 236 \fBlz4\fR is a high-performance real-time compression algorithm that
 237 features significantly faster compression and decompression as well as a
 238 higher compression ratio than the older \fBlzjb\fR compression.
 239 Typically, \fBlz4\fR compression is approximately 50% faster on
 240 compressible data and 200% faster on incompressible data than
 241 \fBlzjb\fR. It is also approximately 80% faster on decompression, while
 242 giving approximately 10% better compression ratio.
 243
 244 When the \fBlz4_compress\fR feature is set to \fBenabled\fR, the
 245 administrator can turn on \fBlz4\fR compression on any dataset on the
 246 pool using the \fBzfs\fR(8) command. Please note that doing so will
 247 immediately activate the \fBlz4_compress\fR feature on the underlying
 248 pool using the \fBzfs\fR(1M) command. Also, all newly written metadata
 249 will be compressed with \fBlz4\fR algorithm. Since this feature is not
 250 read-only compatible, this operation will render the pool unimportable
 251 on systems without support for the \fBlz4_compress\fR feature. Booting
 252 off of \fBlz4\fR-compressed root pools is supported.
 253
 254 This feature becomes \fBactive\fR as soon as it is enabled and will
 255 never return to being \fBenabled\fB.
 256 .RE
 257
 258 .sp
 259 .ne 2
 260 .na
 261 \fB\fBspacemap_histogram\fR\fR
 262 .ad
 263 .RS 4n
 264 .TS
 265 l l .
 266 GUID    com.delphix:spacemap_histogram
 267 READ\-ONLY COMPATIBLE   yes
 268 DEPENDENCIES    none
 269 .TE
 270
 271 This features allows ZFS to maintain more information about how free space
 272 is organized within the pool. If this feature is \fBenabled\fR, ZFS will
 273 set this feature to \fBactive\fR when a new space map object is created or
 274 an existing space map is upgraded to the new format. Once the feature is
 275 \fBactive\fR, it will remain in that state until the pool is destroyed.
 276
 277 .RE
 278
 279 .sp
 280 .ne 2
 281 .na
 282 \fB\fBextensible_dataset\fR\fR
 283 .ad
 284 .RS 4n
 285 .TS
 286 l l .
 287 GUID    com.delphix:extensible_dataset
 288 READ\-ONLY COMPATIBLE   no
 289 DEPENDENCIES    none
 290 .TE
 291
 292 This feature allows more flexible use of internal ZFS data structures,
 293 and exists for other features to depend on.
 294
 295 This feature will be \fBactive\fR when the first dependent feature uses it,
 296 and will be returned to the \fBenabled\fR state when all datasets that use
 297 this feature are destroyed.
 298
 299 .RE
 300
 301 .sp
 302 .ne 2
 303 .na
 304 \fB\fBbookmarks\fR\fR
 305 .ad
 306 .RS 4n
 307 .TS
 308 l l .
 309 GUID    com.delphix:bookmarks
 310 READ\-ONLY COMPATIBLE   yes
 311 DEPENDENCIES    extensible_dataset
 312 .TE
 313
 314 This feature enables use of the \fBzfs bookmark\fR subcommand.
 315
 316 This feature is \fBactive\fR while any bookmarks exist in the pool.
 317 All bookmarks in the pool can be listed by running
 318 \fBzfs list -t bookmark -r \fIpoolname\fR\fR.
 319
 320 .RE
 321
 322 .sp
 323 .ne 2
 324 .na
 325 \fB\fBenabled_txg\fR\fR
 326 .ad
 327 .RS 4n
 328 .TS
 329 l l .
 330 GUID    com.delphix:enabled_txg
 331 READ\-ONLY COMPATIBLE   yes
 332 DEPENDENCIES    none
 333 .TE
 334
 335 Once this feature is enabled ZFS records the transaction group number
 336 in which new features are enabled. This has no user-visible impact,
 337 but other features may depend on this feature.
 338
 339 This feature becomes \fBactive\fR as soon as it is enabled and will
 340 never return to being \fBenabled\fB.
 341
 342 .RE
 343
 344 .sp
 345 .ne 2
 346 .na
 347 \fB\fBhole_birth\fR\fR
 348 .ad
 349 .RS 4n
 350 .TS
 351 l l .
 352 GUID    com.delphix:hole_birth
 353 READ\-ONLY COMPATIBLE   no
 354 DEPENDENCIES    enabled_txg
 355 .TE
 356
 357 This feature improves performance of incremental sends ("zfs send -i")
 358 and receives for objects with many holes. The most common case of
 359 hole-filled objects is zvols.
 360
 361 An incremental send stream from snapshot \fBA\fR to snapshot \fBB\fR
 362 contains information about every block that changed between \fBA\fR and
 363 \fBB\fR. Blocks which did not change between those snapshots can be
 364 identified and omitted from the stream using a piece of metadata called
 365 the 'block birth time', but birth times are not recorded for holes (blocks
 366 filled only with zeroes). Since holes created after \fBA\fR cannot be
 367 distinguished from holes created before \fBA\fR, information about every
 368 hole in the entire filesystem or zvol is included in the send stream.
 369
 370 For workloads where holes are rare this is not a problem. However, when
 371 incrementally replicating filesystems or zvols with many holes (for
 372 example a zvol formatted with another filesystem) a lot of time will
 373 be spent sending and receiving unnecessary information about holes that
 374 already exist on the receiving side.
 375
 376 Once the \fBhole_birth\fR feature has been enabled the block birth times
 377 of all new holes will be recorded. Incremental sends between snapshots
 378 created after this feature is enabled will use this new metadata to avoid
 379 sending information about holes that already exist on the receiving side.
 380
 381 This feature becomes \fBactive\fR as soon as it is enabled and will
 382 never return to being \fBenabled\fB.
 383
 384 .RE
 385
 386 .sp
 387 .ne 2
 388 .na
 389 \fB\fBembedded_data\fR\fR
 390 .ad
 391 .RS 4n
 392 .TS
 393 l l .
 394 GUID    com.delphix:embedded_data
 395 READ\-ONLY COMPATIBLE   no
 396 DEPENDENCIES    none
 397 .TE
 398
 399 This feature improves the performance and compression ratio of
 400 highly-compressible blocks.  Blocks whose contents can compress to 112 bytes
 401 or smaller can take advantage of this feature.
 402
 403 When this feature is enabled, the contents of highly-compressible blocks are
 404 stored in the block "pointer" itself (a misnomer in this case, as it contains
 405 the compressed data, rather than a pointer to its location on disk).  Thus
 406 the space of the block (one sector, typically 512 bytes or 4KB) is saved,
 407 and no additional i/o is needed to read and write the data block.
 408
 409 This feature becomes \fBactive\fR as soon as it is enabled and will
 410 never return to being \fBenabled\fR.
 411
 412 .RE
 413
 414 .sp
 415 .ne 2
 416 .na
 417 \fB\fBlarge_blocks\fR\fR
 418 .ad
 419 .RS 4n
 420 .TS
 421 l l .
 422 GUID    org.open-zfs:large_block
 423 READ\-ONLY COMPATIBLE   no
 424 DEPENDENCIES    extensible_dataset
 425 .TE
 426
 427 The \fBlarge_block\fR feature allows the record size on a dataset to be
 428 set larger than 128KB.
 429
 430 This feature becomes \fBactive\fR once a \fBrecordsize\fR property has been
 431 set larger than 128KB, and will return to being \fBenabled\fR once all
 432 filesystems that have ever had their recordsize larger than 128KB are destroyed.
 433 .RE
 434
 435 .sp
 436 .ne 2
 437 .na
 438 \fB\fBlarge_dnode\fR\fR
 439 .ad
 440 .RS 4n
 441 .TS
 442 l l .
 443 GUID    org.zfsonlinux:large_dnode
 444 READ\-ONLY COMPATIBLE   no
 445 DEPENDENCIES    extensible_dataset
 446 .TE
 447
 448 The \fBlarge_dnode\fR feature allows the size of dnodes in a dataset to be
 449 set larger than 512B.
 450
 451 This feature becomes \fBactive\fR once a dataset contains an object with
 452 a dnode larger than 512B, which occurs as a result of setting the
 453 \fBdnodesize\fR dataset property to a value other than \fBlegacy\fR. The
 454 feature will return to being \fBenabled\fR once all filesystems that
 455 have ever contained a dnode larger than 512B are destroyed. Large dnodes
 456 allow more data to be stored in the bonus buffer, thus potentially
 457 improving performance by avoiding the use of spill blocks.
 458 .RE
 459
 460 \fB\fBsha512\fR\fR
 461 .ad
 462 .RS 4n
 463 .TS
 464 l l .
 465 GUID    org.illumos:sha512
 466 READ\-ONLY COMPATIBLE   no
 467 DEPENDENCIES    extensible_dataset
 468 .TE
 469
 470 This feature enables the use of the SHA-512/256 truncated hash algorithm
 471 (FIPS 180-4) for checksum and dedup. The native 64-bit arithmetic of
 472 SHA-512 provides an approximate 50% performance boost over SHA-256 on
 473 64-bit hardware and is thus a good minimum-change replacement candidate
 474 for systems where hash performance is important, but these systems
 475 cannot for whatever reason utilize the faster \fBskein\fR and
 476 \fBedonr\fR algorithms.
 477
 478 When the \fBsha512\fR feature is set to \fBenabled\fR, the administrator
 479 can turn on the \fBsha512\fR checksum on any dataset using the
 480 \fBzfs set checksum=sha512\fR(1M) command.  This feature becomes
 481 \fBactive\fR once a \fBchecksum\fR property has been set to \fBsha512\fR,
 482 and will return to being \fBenabled\fR once all filesystems that have
 483 ever had their checksum set to \fBsha512\fR are destroyed.
 484
 485 Booting off of pools utilizing SHA-512/256 is supported (provided that
 486 the updated GRUB stage2 module is installed).
 487
 488 .RE
 489
 490 .sp
 491 .ne 2
 492 .na
 493 \fB\fBskein\fR\fR
 494 .ad
 495 .RS 4n
 496 .TS
 497 l l .
 498 GUID    org.illumos:skein
 499 READ\-ONLY COMPATIBLE   no
 500 DEPENDENCIES    extensible_dataset
 501 .TE
 502
 503 This feature enables the use of the Skein hash algorithm for checksum
 504 and dedup. Skein is a high-performance secure hash algorithm that was a
 505 finalist in the NIST SHA-3 competition. It provides a very high security
 506 margin and high performance on 64-bit hardware (80% faster than
 507 SHA-256). This implementation also utilizes the new salted checksumming
 508 functionality in ZFS, which means that the checksum is pre-seeded with a
 509 secret 256-bit random key (stored on the pool) before being fed the data
 510 block to be checksummed. Thus the produced checksums are unique to a
 511 given pool, preventing hash collision attacks on systems with dedup.
 512
 513 When the \fBskein\fR feature is set to \fBenabled\fR, the administrator
 514 can turn on the \fBskein\fR checksum on any dataset using the
 515 \fBzfs set checksum=skein\fR(1M) command.  This feature becomes
 516 \fBactive\fR once a \fBchecksum\fR property has been set to \fBskein\fR,
 517 and will return to being \fBenabled\fR once all filesystems that have
 518 ever had their checksum set to \fBskein\fR are destroyed.
 519
 520 Booting off of pools using \fBskein\fR is \fBNOT\fR supported
 521 -- any attempt to enable \fBskein\fR on a root pool will fail with an
 522 error.
 523
 524 .RE
 525
 526 .sp
 527 .ne 2
 528 .na
 529 \fB\fBedonr\fR\fR
 530 .ad
 531 .RS 4n
 532 .TS
 533 l l .
 534 GUID    org.illumos:edonr
 535 READ\-ONLY COMPATIBLE   no
 536 DEPENDENCIES    extensible_dataset
 537 .TE
 538
 539 This feature enables the use of the Edon-R hash algorithm for checksum,
 540 including for nopwrite (if compression is also enabled, an overwrite of
 541 a block whose checksum matches the data being written will be ignored).
 542 In an abundance of caution, Edon-R can not be used with dedup
 543 (without verification).
 544
 545 Edon-R is a very high-performance hash algorithm that was part
 546 of the NIST SHA-3 competition. It provides extremely high hash
 547 performance (over 350% faster than SHA-256), but was not selected
 548 because of its unsuitability as a general purpose secure hash algorithm.
 549 This implementation utilizes the new salted checksumming functionality
 550 in ZFS, which means that the checksum is pre-seeded with a secret
 551 256-bit random key (stored on the pool) before being fed the data block
 552 to be checksummed. Thus the produced checksums are unique to a given
 553 pool.
 554
 555 When the \fBedonr\fR feature is set to \fBenabled\fR, the administrator
 556 can turn on the \fBedonr\fR checksum on any dataset using the
 557 \fBzfs set checksum=edonr\fR(1M) command.  This feature becomes
 558 \fBactive\fR once a \fBchecksum\fR property has been set to \fBedonr\fR,
 559 and will return to being \fBenabled\fR once all filesystems that have
 560 ever had their checksum set to \fBedonr\fR are destroyed.
 561
 562 Booting off of pools using \fBedonr\fR is \fBNOT\fR supported
 563 -- any attempt to enable \fBedonr\fR on a root pool will fail with an
 564 error.
 565
 566 .SH "SEE ALSO"
 567 \fBzpool\fR(8)