xref: /illumos-gate/usr/src/man/man8/zfs.8 (revision fb876f9607a4e506a871e6b539d75c9644a3566f)
1.\"
2.\" CDDL HEADER START
3.\"
4.\" The contents of this file are subject to the terms of the
5.\" Common Development and Distribution License (the "License").
6.\" You may not use this file except in compliance with the License.
7.\"
8.\" You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
9.\" or http://www.opensolaris.org/os/licensing.
10.\" See the License for the specific language governing permissions
11.\" and limitations under the License.
12.\"
13.\" When distributing Covered Code, include this CDDL HEADER in each
14.\" file and include the License file at usr/src/OPENSOLARIS.LICENSE.
15.\" If applicable, add the following below this CDDL HEADER, with the
16.\" fields enclosed by brackets "[]" replaced with your own identifying
17.\" information: Portions Copyright [yyyy] [name of copyright owner]
18.\"
19.\" CDDL HEADER END
20.\"
21.\"
22.\" Copyright (c) 2009 Sun Microsystems, Inc. All Rights Reserved.
23.\" Copyright 2011 Joshua M. Clulow <josh@sysmgr.org>
24.\" Copyright (c) 2011, 2016 by Delphix. All rights reserved.
25.\" Copyright (c) 2013 by Saso Kiselkov. All rights reserved.
26.\" Copyright (c) 2014 by Adam Stevko. All rights reserved.
27.\" Copyright (c) 2014 Integros [integros.com]
28.\" Copyright 2018 Nexenta Systems, Inc.
29.\" Copyright 2019 Joyent, Inc.
30.\" Copyright (c) 2018 Datto Inc.
31.\" Copyright 2023 RackTop Systems, Inc.
32.\"
33.Dd November 03, 2023
34.Dt ZFS 8
35.Os
36.Sh NAME
37.Nm zfs
38.Nd configures ZFS file systems
39.Sh SYNOPSIS
40.Nm
41.Op Fl \&?
42.Nm
43.Cm create
44.Op Fl Pnpv
45.Oo Fl o Ar property Ns = Ns Ar value Oc Ns ...
46.Ar filesystem
47.Nm
48.Cm create
49.Op Fl Pnpsv
50.Op Fl b Ar blocksize
51.Oo Fl o Ar property Ns = Ns Ar value Oc Ns ...
52.Fl V Ar size Ar volume
53.Nm
54.Cm destroy
55.Op Fl Rfnprv
56.Ar filesystem Ns | Ns Ar volume
57.Nm
58.Cm destroy
59.Op Fl Rdnprv
60.Ar filesystem Ns | Ns Ar volume Ns @ Ns Ar snap Ns
61.Oo % Ns Ar snap Ns Oo , Ns Ar snap Ns Oo % Ns Ar snap Oc Oc Oc Ns ...
62.Nm
63.Cm destroy
64.Ar filesystem Ns | Ns Ar volume Ns # Ns Ar bookmark
65.Nm
66.Cm snapshot
67.Op Fl r
68.Oo Fl o Ar property Ns = Ns value Oc Ns ...
69.Ar filesystem Ns @ Ns Ar snapname Ns | Ns Ar volume Ns @ Ns Ar snapname Ns ...
70.Nm
71.Cm rollback
72.Op Fl Rfr
73.Ar snapshot
74.Nm
75.Cm clone
76.Op Fl p
77.Oo Fl o Ar property Ns = Ns Ar value Oc Ns ...
78.Ar snapshot Ar filesystem Ns | Ns Ar volume
79.Nm
80.Cm promote
81.Ar clone-filesystem
82.Nm
83.Cm rename
84.Op Fl f
85.Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot
86.Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot
87.Nm
88.Cm rename
89.Op Fl fp
90.Ar filesystem Ns | Ns Ar volume
91.Ar filesystem Ns | Ns Ar volume
92.Nm
93.Cm rename
94.Fl r
95.Ar snapshot Ar snapshot
96.Nm
97.Cm list
98.Op Fl r Ns | Ns Fl d Ar depth
99.Op Fl Hp
100.Oo Fl o Ar property Ns Oo , Ns Ar property Oc Ns ... Oc
101.Oo Fl s Ar property Oc Ns ...
102.Oo Fl S Ar property Oc Ns ...
103.Oo Fl t Ar type Ns Oo , Ns Ar type Oc Ns ... Oc
104.Oo Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot Oc Ns ...
105.Nm
106.Cm remap
107.Ar filesystem Ns | Ns Ar volume
108.Nm
109.Cm set
110.Ar property Ns = Ns Ar value Oo Ar property Ns = Ns Ar value Oc Ns ...
111.Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot Ns ...
112.Nm
113.Cm get
114.Op Fl r Ns | Ns Fl d Ar depth
115.Op Fl Hp
116.Oo Fl o Ar field Ns Oo , Ns Ar field Oc Ns ... Oc
117.Oo Fl s Ar source Ns Oo , Ns Ar source Oc Ns ... Oc
118.Oo Fl t Ar type Ns Oo , Ns Ar type Oc Ns ... Oc
119.Cm all | Ar property Ns Oo , Ns Ar property Oc Ns ...
120.Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot Ns | Ns Ar bookmark Ns ...
121.Nm
122.Cm inherit
123.Op Fl rS
124.Ar property Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot Ns ...
125.Nm
126.Cm upgrade
127.Nm
128.Cm upgrade
129.Fl v
130.Nm
131.Cm upgrade
132.Op Fl r
133.Op Fl V Ar version
134.Fl a | Ar filesystem
135.Nm
136.Cm userspace
137.Op Fl Hinp
138.Oo Fl o Ar field Ns Oo , Ns Ar field Oc Ns ... Oc
139.Oo Fl s Ar field Oc Ns ...
140.Oo Fl S Ar field Oc Ns ...
141.Oo Fl t Ar type Ns Oo , Ns Ar type Oc Ns ... Oc
142.Ar filesystem Ns | Ns Ar snapshot
143.Nm
144.Cm groupspace
145.Op Fl Hinp
146.Oo Fl o Ar field Ns Oo , Ns Ar field Oc Ns ... Oc
147.Oo Fl s Ar field Oc Ns ...
148.Oo Fl S Ar field Oc Ns ...
149.Oo Fl t Ar type Ns Oo , Ns Ar type Oc Ns ... Oc
150.Ar filesystem Ns | Ns Ar snapshot
151.Nm
152.Cm projectspace
153.Op Fl Hp
154.Oo Fl o Ar field Ns Oo , Ns Ar field Oc Ns ... Oc
155.Oo Fl s Ar field Oc Ns ...
156.Oo Fl S Ar field Oc Ns ...
157.Ar filesystem Ns | Ns Ar snapshot
158.Nm
159.Cm project
160.Oo Fl d Ns | Ns Fl r Ns Oc
161.Ar file Ns | Ns Ar directory Ns ...
162.Nm
163.Cm project
164.Fl C
165.Oo Fl kr Ns Oc
166.Ar file Ns | Ns Ar directory Ns ...
167.Nm
168.Cm project
169.Fl c
170.Oo Fl 0 Ns Oc
171.Oo Fl d Ns | Ns Fl r Ns Oc
172.Op Fl p Ar id
173.Ar file Ns | Ns Ar directory Ns ...
174.Nm
175.Cm project
176.Op Fl p Ar id
177.Oo Fl rs Ns Oc
178.Ar file Ns | Ns Ar directory Ns ...
179.Nm
180.Cm mount
181.Nm
182.Cm mount
183.Op Fl Olv
184.Op Fl o Ar options
185.Fl a | Ar filesystem
186.Nm
187.Cm unmount
188.Op Fl f
189.Fl a | Ar filesystem Ns | Ns Ar mountpoint
190.Nm
191.Cm share
192.Fl a | Ar filesystem
193.Nm
194.Cm unshare
195.Fl a | Ar filesystem Ns | Ns Ar mountpoint
196.Nm
197.Cm bookmark
198.Ar snapshot bookmark
199.Nm
200.Cm send
201.Op Fl DLPRbcehnpvw
202.Op Oo Fl I Ns | Ns Fl i Oc Ar snapshot
203.Ar snapshot
204.Nm
205.Cm send
206.Op Fl LPcenvw
207.Op Fl i Ar snapshot Ns | Ns Ar bookmark
208.Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot
209.Nm
210.Cm send
211.Op Fl Penv
212.Fl t Ar receive_resume_token
213.Nm
214.Cm receive
215.Op Fl Fhnsuv
216.Op Fl o Sy origin Ns = Ns Ar snapshot
217.Op Fl o Ar property Ns = Ns Ar value
218.Op Fl x Ar property
219.Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot
220.Nm
221.Cm receive
222.Op Fl Fhnsuv
223.Op Fl d Ns | Ns Fl e
224.Op Fl o Sy origin Ns = Ns Ar snapshot
225.Op Fl o Ar property Ns = Ns Ar value
226.Op Fl x Ar property
227.Ar filesystem
228.Nm
229.Cm receive
230.Fl A
231.Ar filesystem Ns | Ns Ar volume
232.Nm
233.Cm allow
234.Ar filesystem Ns | Ns Ar volume
235.Nm
236.Cm allow
237.Op Fl dglu
238.Ar user Ns | Ns Ar group Ns Oo , Ns Ar user Ns | Ns Ar group Oc Ns ...
239.Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
240.Ar setname Oc Ns ...
241.Ar filesystem Ns | Ns Ar volume
242.Nm
243.Cm allow
244.Op Fl dl
245.Fl e Ns | Ns Sy everyone
246.Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
247.Ar setname Oc Ns ...
248.Ar filesystem Ns | Ns Ar volume
249.Nm
250.Cm allow
251.Fl c
252.Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
253.Ar setname Oc Ns ...
254.Ar filesystem Ns | Ns Ar volume
255.Nm
256.Cm allow
257.Fl s No @ Ns Ar setname
258.Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
259.Ar setname Oc Ns ...
260.Ar filesystem Ns | Ns Ar volume
261.Nm
262.Cm unallow
263.Op Fl dglru
264.Ar user Ns | Ns Ar group Ns Oo , Ns Ar user Ns | Ns Ar group Oc Ns ...
265.Oo Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
266.Ar setname Oc Ns ... Oc
267.Ar filesystem Ns | Ns Ar volume
268.Nm
269.Cm unallow
270.Op Fl dlr
271.Fl e Ns | Ns Sy everyone
272.Oo Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
273.Ar setname Oc Ns ... Oc
274.Ar filesystem Ns | Ns Ar volume
275.Nm
276.Cm unallow
277.Op Fl r
278.Fl c
279.Oo Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
280.Ar setname Oc Ns ... Oc
281.Ar filesystem Ns | Ns Ar volume
282.Nm
283.Cm unallow
284.Op Fl r
285.Fl s @ Ns Ar setname
286.Oo Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
287.Ar setname Oc Ns ... Oc
288.Ar filesystem Ns | Ns Ar volume
289.Nm
290.Cm hold
291.Op Fl r
292.Ar tag Ar snapshot Ns ...
293.Nm
294.Cm holds
295.Op Fl r
296.Ar snapshot Ns ...
297.Nm
298.Cm release
299.Op Fl r
300.Ar tag Ar snapshot Ns ...
301.Nm
302.Cm diff
303.Op Fl FHt
304.Ar snapshot Ar snapshot Ns | Ns Ar filesystem
305.Nm
306.Cm program
307.Op Fl jn
308.Op Fl t Ar timeout
309.Op Fl m Ar memory_limit
310.Ar pool script
311.Op Ar arg1 No ...
312.Nm
313.Cm load-key
314.Op Fl rn
315.Op Fl L Ar keylocation
316.Op Fl a Ns | Ns Ar filesystem
317.Nm
318.Cm unload-key
319.Op Fl r
320.Op Fl a Ns | Ns Ar filesystem
321.Nm
322.Cm change-key
323.Op Fl l
324.Op Fl o Sy keylocation Ns = Ns Ar value
325.Op Fl o Sy keyformat Ns = Ns Ar value
326.Op Fl o Sy pbkdf2iters Ns = Ns Ar value
327.Ar filesystem
328.Sh DESCRIPTION
329The
330.Nm
331command configures ZFS datasets within a ZFS storage pool, as described in
332.Xr zpool 8 .
333A dataset is identified by a unique path within the ZFS namespace.
334For example:
335.Bd -literal
336pool/{filesystem,volume,snapshot}
337.Ed
338.Pp
339where the maximum length of a dataset name is
340.Dv MAXNAMELEN
341.Pq 256 bytes
342and the maximum amount of nesting allowed in a path is 50 levels deep.
343.Pp
344A dataset can be one of the following:
345.Bl -tag -width "file system"
346.It Sy file system
347A ZFS dataset of type
348.Sy filesystem
349can be mounted within the standard system namespace and behaves like other file
350systems.
351While ZFS file systems are designed to be POSIX compliant, known issues exist
352that prevent compliance in some cases.
353Applications that depend on standards conformance might fail due to non-standard
354behavior when checking file system free space.
355.It Sy volume
356A logical volume exported as a raw or block device.
357This type of dataset should only be used under special circumstances.
358File systems are typically used in most environments.
359.It Sy snapshot
360A read-only version of a file system or volume at a given point in time.
361It is specified as
362.Ar filesystem Ns @ Ns Ar name
363or
364.Ar volume Ns @ Ns Ar name .
365.El
366.Ss ZFS File System Hierarchy
367A ZFS storage pool is a logical collection of devices that provide space for
368datasets.
369A storage pool is also the root of the ZFS file system hierarchy.
370.Pp
371The root of the pool can be accessed as a file system, such as mounting and
372unmounting, taking snapshots, and setting properties.
373The physical storage characteristics, however, are managed by the
374.Xr zpool 8
375command.
376.Pp
377See
378.Xr zpool 8
379for more information on creating and administering pools.
380.Ss Snapshots
381A snapshot is a read-only copy of a file system or volume.
382Snapshots can be created extremely quickly, and initially consume no additional
383space within the pool.
384As data within the active dataset changes, the snapshot consumes more data than
385would otherwise be shared with the active dataset.
386.Pp
387Snapshots can have arbitrary names.
388Snapshots of volumes can be cloned or rolled back, but cannot be accessed
389independently.
390.Pp
391File system snapshots can be accessed under the
392.Pa .zfs/snapshot
393directory in the root of the file system.
394Snapshots are automatically mounted on demand and may be unmounted at regular
395intervals.
396The visibility of the
397.Pa .zfs
398directory can be controlled by the
399.Sy snapdir
400property.
401.Ss Clones
402A clone is a writable volume or file system whose initial contents are the same
403as another dataset.
404As with snapshots, creating a clone is nearly instantaneous, and initially
405consumes no additional space.
406.Pp
407Clones can only be created from a snapshot.
408When a snapshot is cloned, it creates an implicit dependency between the parent
409and child.
410Even though the clone is created somewhere else in the dataset hierarchy, the
411original snapshot cannot be destroyed as long as a clone exists.
412The
413.Sy origin
414property exposes this dependency, and the
415.Cm destroy
416command lists any such dependencies, if they exist.
417.Pp
418The clone parent-child dependency relationship can be reversed by using the
419.Cm promote
420subcommand.
421This causes the
422.Qq origin
423file system to become a clone of the specified file system, which makes it
424possible to destroy the file system that the clone was created from.
425.Ss "Mount Points"
426Creating a ZFS file system is a simple operation, so the number of file systems
427per system is likely to be numerous.
428To cope with this, ZFS automatically manages mounting and unmounting file
429systems without the need to edit the
430.Pa /etc/vfstab
431file.
432All automatically managed file systems are mounted by ZFS at boot time.
433.Pp
434By default, file systems are mounted under
435.Pa /path ,
436where
437.Ar path
438is the name of the file system in the ZFS namespace.
439Directories are created and destroyed as needed.
440.Pp
441A file system can also have a mount point set in the
442.Sy mountpoint
443property.
444This directory is created as needed, and ZFS automatically mounts the file
445system when the
446.Nm zfs Cm mount Fl a
447command is invoked
448.Po without editing
449.Pa /etc/vfstab
450.Pc .
451The
452.Sy mountpoint
453property can be inherited, so if
454.Em pool/home
455has a mount point of
456.Pa /export/stuff ,
457then
458.Em pool/home/user
459automatically inherits a mount point of
460.Pa /export/stuff/user .
461.Pp
462A file system
463.Sy mountpoint
464property of
465.Sy none
466prevents the file system from being mounted.
467.Pp
468If needed, ZFS file systems can also be managed with traditional tools
469.Po
470.Nm mount ,
471.Nm umount ,
472.Pa /etc/vfstab
473.Pc .
474If a file system's mount point is set to
475.Sy legacy ,
476ZFS makes no attempt to manage the file system, and the administrator is
477responsible for mounting and unmounting the file system.
478.Ss "Zones"
479A ZFS file system can be added to a non-global zone by using the
480.Nm zonecfg Cm add Sy fs
481subcommand.
482A ZFS file system that is added to a non-global zone must have its
483.Sy mountpoint
484property set to
485.Sy legacy .
486.Pp
487The physical properties of an added file system are controlled by the global
488administrator.
489However, the zone administrator can create, modify, or destroy files within the
490added file system, depending on how the file system is mounted.
491.Pp
492A dataset can also be delegated to a non-global zone by using the
493.Nm zonecfg Cm add Sy dataset
494subcommand.
495You cannot delegate a dataset to one zone and the children of the same dataset
496to another zone.
497The zone administrator can change properties of the dataset or any of its
498children.
499However, the
500.Sy quota ,
501.Sy filesystem_limit
502and
503.Sy snapshot_limit
504properties of the delegated dataset can be modified only by the global
505administrator.
506.Pp
507A ZFS volume can be added as a device to a non-global zone by using the
508.Nm zonecfg Cm add Sy device
509subcommand.
510However, its physical properties can be modified only by the global
511administrator.
512.Pp
513For more information about
514.Nm zonecfg
515syntax, see
516.Xr zonecfg 8 .
517.Pp
518After a dataset is delegated to a non-global zone, the
519.Sy zoned
520property is automatically set.
521A zoned file system cannot be mounted in the global zone, since the zone
522administrator might have to set the mount point to an unacceptable value.
523.Pp
524The global administrator can forcibly clear the
525.Sy zoned
526property, though this should be done with extreme care.
527The global administrator should verify that all the mount points are acceptable
528before clearing the property.
529.Ss Native Properties
530Properties are divided into two types, native properties and user-defined
531.Po or
532.Qq user
533.Pc
534properties.
535Native properties either export internal statistics or control ZFS behavior.
536In addition, native properties are either editable or read-only.
537User properties have no effect on ZFS behavior, but you can use them to annotate
538datasets in a way that is meaningful in your environment.
539For more information about user properties, see the
540.Sx User Properties
541section, below.
542.Pp
543Every dataset has a set of properties that export statistics about the dataset
544as well as control various behaviors.
545Properties are inherited from the parent unless overridden by the child.
546Some properties apply only to certain types of datasets
547.Pq file systems, volumes, or snapshots .
548.Pp
549The values of numeric properties can be specified using human-readable suffixes
550.Po for example,
551.Sy k ,
552.Sy KB ,
553.Sy M ,
554.Sy Gb ,
555and so forth, up to
556.Sy Z
557for zettabyte
558.Pc .
559The following are all valid
560.Pq and equal
561specifications:
562.Li 1536M, 1.5g, 1.50GB .
563.Pp
564The values of non-numeric properties are case sensitive and must be lowercase,
565except for
566.Sy mountpoint ,
567.Sy sharenfs ,
568and
569.Sy sharesmb .
570.Pp
571The following native properties consist of read-only statistics about the
572dataset.
573These properties can be neither set, nor inherited.
574Native properties apply to all dataset types unless otherwise noted.
575.Bl -tag -width "usedbyrefreservation"
576.It Sy available
577The amount of space available to the dataset and all its children, assuming that
578there is no other activity in the pool.
579Because space is shared within a pool, availability can be limited by any number
580of factors, including physical pool size, quotas, reservations, or other
581datasets within the pool.
582.Pp
583This property can also be referred to by its shortened column name,
584.Sy avail .
585.It Sy compressratio
586For non-snapshots, the compression ratio achieved for the
587.Sy used
588space of this dataset, expressed as a multiplier.
589The
590.Sy used
591property includes descendant datasets, and, for clones, does not include the
592space shared with the origin snapshot.
593For snapshots, the
594.Sy compressratio
595is the same as the
596.Sy refcompressratio
597property.
598Compression can be turned on by running:
599.Nm zfs Cm set Sy compression Ns = Ns Sy on Ar dataset .
600The default value is
601.Sy off .
602.It Sy createtxg
603The transaction group (txg) in which the dataset was created.
604Bookmarks have the same
605.Sy createtxg
606as the snapshot they are initially tied to.
607This property is suitable for ordering a list of snapshots,
608e.g. for incremental send and receive.
609.It Sy creation
610The time this dataset was created.
611.It Sy clones
612For snapshots, this property is a comma-separated list of filesystems or volumes
613which are clones of this snapshot.
614The clones'
615.Sy origin
616property is this snapshot.
617If the
618.Sy clones
619property is not empty, then this snapshot can not be destroyed
620.Po even with the
621.Fl r
622or
623.Fl f
624options
625.Pc .
626.It Sy defer_destroy
627This property is
628.Sy on
629if the snapshot has been marked for deferred destroy by using the
630.Nm zfs Cm destroy Fl d
631command.
632Otherwise, the property is
633.Sy off .
634.It Sy encryptionroot
635For encrypted datasets, indicates where the dataset is currently inheriting its
636encryption key from.
637Loading or unloading a key for the
638.Sy encryptionroot
639will implicitly load / unload the key for any inheriting datasets
640.Po see
641.Nm zfs Cm load-key
642and
643.Nm zfs Cm unload-key
644.Pc .
645Clones will always share an encryption key with their origin.
646See the
647.Sy Encryption
648section for details.
649.It Sy filesystem_count
650The total number of filesystems and volumes that exist under this location in
651the dataset tree.
652This value is only available when a
653.Sy filesystem_limit
654has been set somewhere in the tree under which the dataset resides.
655.It Sy guid
656The 64 bit GUID of this dataset or bookmark which does not change over its
657entire lifetime.
658When a snapshot is sent to another pool, the received snapshot has the same
659GUID.
660Thus, the
661.Sy guid
662is suitable to identify a snapshot across pools.
663.It Sy keystatus
664Indicates if an encryption key is currently loaded into ZFS.
665The possible values are
666.Sy none , available ,
667and
668.Sy unavailable .
669See
670.Nm Cm load-key
671and
672.Nm Cm unload-key .
673.It Sy logicalreferenced
674The amount of space that is
675.Qq logically
676accessible by this dataset.
677See the
678.Sy referenced
679property.
680The logical space ignores the effect of the
681.Sy compression
682and
683.Sy copies
684properties, giving a quantity closer to the amount of data that applications
685see.
686However, it does include space consumed by metadata.
687.Pp
688This property can also be referred to by its shortened column name,
689.Sy lrefer .
690.It Sy logicalused
691The amount of space that is
692.Qq logically
693consumed by this dataset and all its descendents.
694See the
695.Sy used
696property.
697The logical space ignores the effect of the
698.Sy compression
699and
700.Sy copies
701properties, giving a quantity closer to the amount of data that applications
702see.
703However, it does include space consumed by metadata.
704.Pp
705This property can also be referred to by its shortened column name,
706.Sy lused .
707.It Sy mounted
708For file systems, indicates whether the file system is currently mounted.
709This property can be either
710.Sy yes
711or
712.Sy no .
713.It Sy origin
714For cloned file systems or volumes, the snapshot from which the clone was
715created.
716See also the
717.Sy clones
718property.
719.It Sy receive_resume_token
720For filesystems or volumes which have saved partially-completed state from
721.Sy zfs receive -s ,
722this opaque token can be provided to
723.Sy zfs send -t
724to resume and complete the
725.Sy zfs receive .
726.It Sy referenced
727The amount of data that is accessible by this dataset, which may or may not be
728shared with other datasets in the pool.
729When a snapshot or clone is created, it initially references the same amount of
730space as the file system or snapshot it was created from, since its contents are
731identical.
732.Pp
733This property can also be referred to by its shortened column name,
734.Sy refer .
735.It Sy refcompressratio
736The compression ratio achieved for the
737.Sy referenced
738space of this dataset, expressed as a multiplier.
739See also the
740.Sy compressratio
741property.
742.It Sy snapshot_count
743The total number of snapshots that exist under this location in the dataset
744tree.
745This value is only available when a
746.Sy snapshot_limit
747has been set somewhere in the tree under which the dataset resides.
748.It Sy type
749The type of dataset:
750.Sy filesystem ,
751.Sy volume ,
752or
753.Sy snapshot .
754.It Sy used
755The amount of space consumed by this dataset and all its descendents.
756This is the value that is checked against this dataset's quota and reservation.
757The space used does not include this dataset's reservation, but does take into
758account the reservations of any descendent datasets.
759The amount of space that a dataset consumes from its parent, as well as the
760amount of space that is freed if this dataset is recursively destroyed, is the
761greater of its space used and its reservation.
762.Pp
763The used space of a snapshot
764.Po see the
765.Sx Snapshots
766section
767.Pc
768is space that is referenced exclusively by this snapshot.
769If this snapshot is destroyed, the amount of
770.Sy used
771space will be freed.
772Space that is shared by multiple snapshots isn't accounted for in this metric.
773When a snapshot is destroyed, space that was previously shared with this
774snapshot can become unique to snapshots adjacent to it, thus changing the used
775space of those snapshots.
776The used space of the latest snapshot can also be affected by changes in the
777file system.
778Note that the
779.Sy used
780space of a snapshot is a subset of the
781.Sy written
782space of the snapshot.
783.Pp
784The amount of space used, available, or referenced does not take into account
785pending changes.
786Pending changes are generally accounted for within a few seconds.
787Committing a change to a disk using
788.Xr fsync 3C
789or
790.Dv O_SYNC
791does not necessarily guarantee that the space usage information is updated
792immediately.
793.It Sy usedby*
794The
795.Sy usedby*
796properties decompose the
797.Sy used
798properties into the various reasons that space is used.
799Specifically,
800.Sy used No =
801.Sy usedbychildren No +
802.Sy usedbydataset No +
803.Sy usedbyrefreservation No +
804.Sy usedbysnapshots .
805These properties are only available for datasets created on
806.Nm zpool
807.Qo version 13 Qc
808pools.
809.It Sy usedbychildren
810The amount of space used by children of this dataset, which would be freed if
811all the dataset's children were destroyed.
812.It Sy usedbydataset
813The amount of space used by this dataset itself, which would be freed if the
814dataset were destroyed
815.Po after first removing any
816.Sy refreservation
817and destroying any necessary snapshots or descendents
818.Pc .
819.It Sy usedbyrefreservation
820The amount of space used by a
821.Sy refreservation
822set on this dataset, which would be freed if the
823.Sy refreservation
824was removed.
825.It Sy usedbysnapshots
826The amount of space consumed by snapshots of this dataset.
827In particular, it is the amount of space that would be freed if all of this
828dataset's snapshots were destroyed.
829Note that this is not simply the sum of the snapshots'
830.Sy used
831properties because space can be shared by multiple snapshots.
832.It Sy userused Ns @ Ns Em user
833The amount of space consumed by the specified user in this dataset.
834Space is charged to the owner of each file, as displayed by
835.Nm ls Fl l .
836The amount of space charged is displayed by
837.Nm du
838and
839.Nm ls Fl s .
840See the
841.Nm zfs Cm userspace
842subcommand for more information.
843.Pp
844Unprivileged users can access only their own space usage.
845The root user, or a user who has been granted the
846.Sy userused
847privilege with
848.Nm zfs Cm allow ,
849can access everyone's usage.
850.Pp
851The
852.Sy userused Ns @ Ns Em ...
853properties are not displayed by
854.Nm zfs Cm get Sy all .
855The user's name must be appended after the @ symbol, using one of the following
856forms:
857.Bl -bullet -width ""
858.It
859.Em POSIX name
860.Po for example,
861.Sy joe
862.Pc
863.It
864.Em POSIX numeric ID
865.Po for example,
866.Sy 789
867.Pc
868.It
869.Em SID name
870.Po for example,
871.Sy joe.smith@mydomain
872.Pc
873.It
874.Em SID numeric ID
875.Po for example,
876.Sy S-1-123-456-789
877.Pc
878.El
879.It Sy userobjused Ns @ Ns Em user
880The
881.Sy userobjused
882property is similar to
883.Sy userused
884but instead it counts the number of objects consumed by a user.
885This property counts all objects allocated on behalf of the user, it may
886differ from the results of system tools such as
887.Nm df Fl i .
888.Pp
889When the property
890.Sy xattr=on
891is set on a file system additional objects will be created per-file to store
892extended attributes.
893These additional objects are reflected in the
894.Sy userobjused
895value and are counted against the user's
896.Sy userobjquota .
897.It Sy userrefs
898This property is set to the number of user holds on this snapshot.
899User holds are set by using the
900.Nm zfs Cm hold
901command.
902.It Sy groupused Ns @ Ns Em group
903The amount of space consumed by the specified group in this dataset.
904Space is charged to the group of each file, as displayed by
905.Nm ls Fl l .
906See the
907.Sy userused Ns @ Ns Em user
908property for more information.
909.Pp
910Unprivileged users can only access their own groups' space usage.
911The root user, or a user who has been granted the
912.Sy groupused
913privilege with
914.Nm zfs Cm allow ,
915can access all groups' usage.
916.It Sy groupobjused Ns @ Ns Em group
917The number of objects consumed by the specified group in this dataset.
918Multiple objects may be charged to the group for each file when extended
919attributes are in use.
920See the
921.Sy userobjused Ns @ Ns Em user
922property for more information.
923.Pp
924Unprivileged users can only access their own groups' space usage.
925The root user, or a user who has been granted the
926.Sy groupobjused
927privilege with
928.Nm zfs Cm allow ,
929can access all groups' usage.
930.It Sy projectused Ns @ Ns Em project
931The amount of space consumed by the specified project in this dataset.
932Project is identified via the project identifier (ID) that is object-based
933numeral attribute.
934An object can inherit the project ID from its parent object (if the
935parent has the flag of inherit project ID that can be set and changed via
936.Nm zfs project Fl s )
937when being created.
938The privileged user can set and change object's project
939ID via
940.Nm zfs project Fl s
941anytime.
942Space is charged to the project of each file, as displayed by
943.Nm zfs project .
944See the
945.Sy userused Ns @ Ns Em user
946property for more information.
947.Pp
948The root user, or a user who has been granted the
949.Sy projectused
950privilege with
951.Nm zfs allow ,
952can access all projects' usage.
953.It Sy projectobjused Ns @ Ns Em project
954The
955.Sy projectobjused
956is similar to
957.Sy projectused
958but instead it counts the number of objects consumed by project.
959When the property
960.Sy xattr=on
961is set on a fileset, ZFS will create additional objects per-file to store
962extended attributes.
963These additional objects are reflected in the
964.Sy projectobjused
965value and are counted against the project's
966.Sy projectobjquota .
967See the
968.Sy userobjused Ns @ Ns Em user
969property for more information.
970.Pp
971The root user, or a user who has been granted the
972.Sy projectobjused
973privilege with
974.Nm zfs allow ,
975can access all projects' objects usage.
976.It Sy volblocksize
977For volumes, specifies the block size of the volume.
978The
979.Sy blocksize
980cannot be changed once the volume has been written, so it should be set at
981volume creation time.
982The default
983.Sy blocksize
984for volumes is 8 Kbytes.
985Any power of 2 from 512 bytes to 128 Kbytes is valid.
986.Pp
987This property can also be referred to by its shortened column name,
988.Sy volblock .
989.It Sy written
990The amount of space
991.Sy referenced
992by this dataset, that was written since the previous snapshot
993.Pq i.e. that is not referenced by the previous snapshot .
994.It Sy written Ns @ Ns Em snapshot
995The amount of
996.Sy referenced
997space written to this dataset since the specified snapshot.
998This is the space that is referenced by this dataset but was not referenced by
999the specified snapshot.
1000.Pp
1001The
1002.Em snapshot
1003may be specified as a short snapshot name
1004.Po just the part after the
1005.Sy @
1006.Pc ,
1007in which case it will be interpreted as a snapshot in the same filesystem as
1008this dataset.
1009The
1010.Em snapshot
1011may be a full snapshot name
1012.Po Em filesystem Ns @ Ns Em snapshot Pc ,
1013which for clones may be a snapshot in the origin's filesystem
1014.Pq or the origin of the origin's filesystem, etc.
1015.El
1016.Pp
1017The following native properties can be used to change the behavior of a ZFS
1018dataset.
1019.Bl -tag -width ""
1020.It Xo
1021.Sy aclinherit Ns = Ns Sy discard Ns | Ns Sy noallow Ns | Ns
1022.Sy restricted Ns | Ns Sy passthrough Ns | Ns Sy passthrough-x
1023.Xc
1024Controls how ACEs are inherited when files and directories are created.
1025.Bl -tag -width "passthrough-x"
1026.It Sy discard
1027does not inherit any ACEs.
1028.It Sy noallow
1029only inherits inheritable ACEs that specify
1030.Qq deny
1031permissions.
1032.It Sy restricted
1033default, removes the
1034.Sy write_acl
1035and
1036.Sy write_owner
1037permissions when the ACE is inherited.
1038.It Sy passthrough
1039inherits all inheritable ACEs without any modifications.
1040.It Sy passthrough-x
1041same meaning as
1042.Sy passthrough ,
1043except that the
1044.Sy owner@ ,
1045.Sy group@ ,
1046and
1047.Sy everyone@
1048ACEs inherit the execute permission only if the file creation mode also requests
1049the execute bit.
1050.El
1051.Pp
1052When the property value is set to
1053.Sy passthrough ,
1054files are created with a mode determined by the inheritable ACEs.
1055If no inheritable ACEs exist that affect the mode, then the mode is set in
1056accordance to the requested mode from the application.
1057.It Xo
1058.Sy aclmode Ns = Ns Sy discard Ns | Ns Sy groupmask Ns | Ns
1059.Sy passthrough Ns | Ns Sy restricted
1060.Xc
1061Controls how an ACL is modified during
1062.Xr chmod 2
1063and how inherited ACEs are modified by the file creation mode.
1064.Bl -tag -width "passthrough"
1065.It Sy discard
1066default, deletes all ACEs except for those representing the mode of the file or
1067directory requested by
1068.Xr chmod 2 .
1069.It Sy groupmask
1070reduces permissions granted by all
1071.Sy ALLOW
1072entries found in the ACL such that they are no greater than the group
1073permissions specified by the mode.
1074.It Sy passthrough
1075indicates that no changes are made to the ACL other than creating or updating
1076the necessary ACEs to represent the new mode of the file or directory.
1077.It Sy restricted
1078causes the
1079.Xr chmod 2
1080operation to return an error when used on any file or directory which has a
1081non-trivial ACL, with entries in addition to those that represent the mode.
1082.El
1083.Pp
1084.Xr chmod 2
1085is required to change the set user ID, set group ID, or sticky bit on a file or
1086directory, as they do not have equivalent ACEs.
1087In order to use
1088.Xr chmod 2
1089on a file or directory with a non-trivial ACL when
1090.Sy aclmode
1091is set to
1092.Sy restricted ,
1093you must first remove all ACEs except for those that represent the current mode.
1094.It Sy aclimplicit  Ns = Ns Sy on Ns | Ns Sy off
1095Controls whether the owner of an object has "implicit owner rights".
1096If this property is set to
1097.Sy on ,
1098then the owner of an object can always
1099.Xr chmod 2
1100as is expected with traditional
1101.Em POSIX
1102file permissions.
1103If this property is set to
1104.Sy off ,
1105then the owner may only
1106.Xr chmod 2
1107an object where the ACL grants
1108.Sy write_acl
1109to the user attempting the action.
1110The default value is
1111.Sy on .
1112Note that
1113.Sy aclimplicit  Ns = Ns Sy off
1114is only fully effective with
1115.Sy aclmode  Ns = Ns Sy passthrough
1116and
1117.Sy aclinherit  Ns = Ns Sy passthrough
1118because in other configurations, objects can end up having
1119.Sy write_acl
1120granted to the object owner.
1121.It Sy atime Ns = Ns Sy on Ns | Ns Sy off
1122Controls whether the access time for files is updated when they are read.
1123Turning this property off avoids producing write traffic when reading files and
1124can result in significant performance gains, though it might confuse mailers
1125and other similar utilities.
1126The default value is
1127.Sy on .
1128.It Sy canmount Ns = Ns Sy on Ns | Ns Sy off Ns | Ns Sy noauto
1129If this property is set to
1130.Sy off ,
1131the file system cannot be mounted, and is ignored by
1132.Nm zfs Cm mount Fl a .
1133Setting this property to
1134.Sy off
1135is similar to setting the
1136.Sy mountpoint
1137property to
1138.Sy none ,
1139except that the dataset still has a normal
1140.Sy mountpoint
1141property, which can be inherited.
1142Setting this property to
1143.Sy off
1144allows datasets to be used solely as a mechanism to inherit properties.
1145One example of setting
1146.Sy canmount Ns = Ns Sy off
1147is to have two datasets with the same
1148.Sy mountpoint ,
1149so that the children of both datasets appear in the same directory, but might
1150have different inherited characteristics.
1151.Pp
1152When set to
1153.Sy noauto ,
1154a dataset can only be mounted and unmounted explicitly.
1155The dataset is not mounted automatically when the dataset is created or
1156imported, nor is it mounted by the
1157.Nm zfs Cm mount Fl a
1158command or unmounted by the
1159.Nm zfs Cm unmount Fl a
1160command.
1161.Pp
1162This property is not inherited.
1163.It Xo
1164.Sy checksum Ns = Ns Sy on Ns | Ns Sy off Ns | Ns Sy fletcher2 Ns | Ns
1165.Sy fletcher4 Ns | Ns Sy sha256 Ns | Ns Sy noparity Ns | Ns
1166.Sy sha512 Ns | Ns Sy skein Ns | Ns Sy edonr
1167.Xc
1168Controls the checksum used to verify data integrity.
1169The default value is
1170.Sy on ,
1171which automatically selects an appropriate algorithm
1172.Po currently,
1173.Sy fletcher4 ,
1174but this may change in future releases
1175.Pc .
1176The value
1177.Sy off
1178disables integrity checking on user data.
1179The value
1180.Sy noparity
1181not only disables integrity but also disables maintaining parity for user data.
1182This setting is used internally by a dump device residing on a RAID-Z pool and
1183should not be used by any other dataset.
1184Disabling checksums is
1185.Sy NOT
1186a recommended practice.
1187.Pp
1188The
1189.Sy sha512 ,
1190.Sy skein ,
1191and
1192.Sy edonr
1193checksum algorithms require enabling the appropriate features on the pool.
1194Please see
1195.Xr zpool-features 7
1196for more information on these algorithms.
1197.Pp
1198Changing this property affects only newly-written data.
1199.It Xo
1200.Sy compression Ns = Ns Sy on Ns | Ns Sy off Ns | Ns Sy gzip Ns | Ns
1201.Sy gzip- Ns Em N Ns | Ns Sy lz4 Ns | Ns Sy lzjb Ns | Ns Sy zle
1202.Xc
1203Controls the compression algorithm used for this dataset.
1204.Pp
1205Setting compression to
1206.Sy on
1207indicates that the current default compression algorithm should be used.
1208The default balances compression and decompression speed, with compression ratio
1209and is expected to work well on a wide variety of workloads.
1210Unlike all other settings for this property,
1211.Sy on
1212does not select a fixed compression type.
1213As new compression algorithms are added to ZFS and enabled on a pool, the
1214default compression algorithm may change.
1215The current default compression algorithm is either
1216.Sy lzjb
1217or, if the
1218.Sy lz4_compress
1219feature is enabled,
1220.Sy lz4 .
1221.Pp
1222The
1223.Sy lz4
1224compression algorithm is a high-performance replacement for the
1225.Sy lzjb
1226algorithm.
1227It features significantly faster compression and decompression, as well as a
1228moderately higher compression ratio than
1229.Sy lzjb ,
1230but can only be used on pools with the
1231.Sy lz4_compress
1232feature set to
1233.Sy enabled .
1234See
1235.Xr zpool-features 7
1236for details on ZFS feature flags and the
1237.Sy lz4_compress
1238feature.
1239.Pp
1240The
1241.Sy lzjb
1242compression algorithm is optimized for performance while providing decent data
1243compression.
1244.Pp
1245The
1246.Sy gzip
1247compression algorithm uses the same compression as the
1248.Xr gzip 1
1249command.
1250You can specify the
1251.Sy gzip
1252level by using the value
1253.Sy gzip- Ns Em N ,
1254where
1255.Em N
1256is an integer from 1
1257.Pq fastest
1258to 9
1259.Pq best compression ratio .
1260Currently,
1261.Sy gzip
1262is equivalent to
1263.Sy gzip-6
1264.Po which is also the default for
1265.Xr gzip 1
1266.Pc .
1267.Pp
1268The
1269.Sy zle
1270compression algorithm compresses runs of zeros.
1271.Pp
1272This property can also be referred to by its shortened column name
1273.Sy compress .
1274Changing this property affects only newly-written data.
1275.It Sy copies Ns = Ns Sy 1 Ns | Ns Sy 2 Ns | Ns Sy 3
1276Controls the number of copies of data stored for this dataset.
1277These copies are in addition to any redundancy provided by the pool, for
1278example, mirroring or RAID-Z.
1279The copies are stored on different disks, if possible.
1280The space used by multiple copies is charged to the associated file and dataset,
1281changing the
1282.Sy used
1283property and counting against quotas and reservations.
1284.Pp
1285Changing this property only affects newly-written data.
1286Therefore, set this property at file system creation time by using the
1287.Fl o Sy copies Ns = Ns Ar N
1288option.
1289.It Sy devices Ns = Ns Sy on Ns | Ns Sy off
1290Controls whether device nodes can be opened on this file system.
1291The default value is
1292.Sy on .
1293.It Xo
1294.Sy encryption Ns = Ns Sy on Ns | Ns Sy off Ns | Ns Sy aes-128-ccm Ns | Ns
1295.Sy aes-192-ccm Ns | Ns Sy aes-256-ccm Ns | Ns Sy aes-128-gcm Ns | Ns
1296.Sy aes-192-gcm Ns | Ns Sy aes-256-gcm
1297.Xc
1298Controls the encryption cipher suite
1299.Pq block cipher, key length, and mode
1300used for this dataset.
1301Requires the encryption feature to be enabled on the pool.
1302Requires a
1303.Sy keyformat
1304to be set at dataset creation time.
1305.Pp
1306Selecting
1307.Sy encryption Ns = Ns Sy on
1308when creating a dataset indicates that the default encryption suite will be
1309selected, which is currently
1310.Sy aes-256-ccm .
1311In order to provide consistent data protection, encryption must be specified at
1312dataset creation time and it cannot be changed afterwards.
1313.Pp
1314For more details and caveats about encryption see the
1315.Sx Encryption
1316section.
1317.It Sy keyformat Ns = Ns Sy raw Ns | Ns Sy hex Ns | Ns Sy passphrase
1318Controls what format the user's encryption key will be provided as.
1319This property is only set for encrypted datasets which are encryption roots.
1320.Pp
1321Raw keys and hex keys must be 32 bytes long
1322.Pq regardless of the chosen encryption suite
1323and must be randomly generated.
1324A raw key can be generated with the following command:
1325.Bd -literal
1326# dd if=/dev/urandom of=/path/to/output/key bs=32 count=1
1327.Ed
1328.Pp
1329Passphrases must be between 8 and 512 bytes long and will be processed through
1330PBKDF2 before being used
1331.Po see the
1332.Nm pbkdf2iters
1333property
1334.Pc .
1335Even though the encryption suite cannot be changed after dataset creation, the
1336keyformat can be with
1337.Nm Cm change-key .
1338.It Sy keylocation Ns = Ns Sy prompt Ns | Ns Ar file://<absolute file path>
1339Controls where the user's encryption key will be loaded from by default for
1340commands such as
1341.Nm Cm load-key
1342and
1343.Nm Cm mount Fl l .
1344This property is only set for encrypted datasets which are encryption roots.
1345If unspecified, the default is
1346.Sy prompt .
1347.Pp
1348Even though the encryption suite cannot be changed after dataset creation, the
1349keylocation can be with either
1350.Nm Cm set
1351or
1352.Nm Cm change-key .
1353If
1354.Sy prompt
1355is selected ZFS will ask for the key at the command prompt when
1356it is required to access the encrypted data
1357.Po see
1358.Nm Cm load-key
1359.Pc .
1360This setting will also allow the key to be passed in via STDIN, but users
1361should be careful not to place keys which should be kept secret on the
1362command line.
1363If a file URI is selected, the key will be loaded from the specified absolute
1364file path.
1365.It Sy exec Ns = Ns Sy on Ns | Ns Sy off
1366Controls whether processes can be executed from within this file system.
1367The default value is
1368.Sy on .
1369.It Sy filesystem_limit Ns = Ns Em count Ns | Ns Sy none
1370Limits the number of filesystems and volumes that can exist under this point in
1371the dataset tree.
1372The limit is not enforced if the user is allowed to change the limit.
1373Setting a
1374.Sy filesystem_limit
1375to
1376.Sy on
1377a descendent of a filesystem that already has a
1378.Sy filesystem_limit
1379does not override the ancestor's
1380.Sy filesystem_limit ,
1381but rather imposes an additional limit.
1382This feature must be enabled to be used
1383.Po see
1384.Xr zpool-features 7
1385.Pc .
1386.It Sy special_small_blocks Ns = Ns Em size
1387This value represents the threshold block size for including small file
1388blocks into the special allocation class.
1389Blocks smaller than or equal to this value will be assigned to the special
1390allocation class while greater blocks will be assigned to the regular class.
1391Valid values are zero or a power of two from 512B up to 128K.
1392The default size is 0 which means no small file blocks will be allocated in
1393the special class.
1394.Pp
1395Before setting this property, a special class vdev must be added to the
1396pool.
1397See
1398.Xr zpool 8
1399for more details on the special allocation class.
1400.It Sy mountpoint Ns = Ns Pa path Ns | Ns Sy none Ns | Ns Sy legacy
1401Controls the mount point used for this file system.
1402See the
1403.Sx Mount Points
1404section for more information on how this property is used.
1405.Pp
1406When the
1407.Sy mountpoint
1408property is changed for a file system, the file system and any children that
1409inherit the mount point are unmounted.
1410If the new value is
1411.Sy legacy ,
1412then they remain unmounted.
1413Otherwise, they are automatically remounted in the new location if the property
1414was previously
1415.Sy legacy
1416or
1417.Sy none ,
1418or if they were mounted before the property was changed.
1419In addition, any shared file systems are unshared and shared in the new
1420location.
1421.It Sy nbmand Ns = Ns Sy on Ns | Ns Sy off
1422Controls whether the file system should be mounted with
1423.Sy nbmand
1424.Pq Non Blocking mandatory locks .
1425This is used for SMB clients.
1426Changes to this property only take effect when the file system is umounted and
1427remounted.
1428See
1429.Xr mount 8
1430for more information on
1431.Sy nbmand
1432mounts.
1433.It Sy pbkdf2iters Ns = Ns Ar iterations
1434Controls the number of PBKDF2 iterations that a
1435.Sy passphrase
1436encryption key should be run through when processing it into an encryption key.
1437This property is only defined when encryption is enabled and a keyformat of
1438.Sy passphrase
1439is selected.
1440The goal of PBKDF2 is to significantly increase the computational difficulty
1441needed to brute force a user's passphrase.
1442This is accomplished by forcing the attacker to run each passphrase through a
1443computationally expensive hashing function many times before they arrive at the
1444resulting key.
1445A user who actually knows the passphrase will only have to pay this cost once.
1446As CPUs become better at processing, this number should be raised to ensure that
1447a brute force attack is still not possible.
1448The current default is 350000 and the minimum is 100000.
1449This property may be changed with
1450.Nm Cm change-key .
1451.It Sy primarycache Ns = Ns Sy all Ns | Ns Sy none Ns | Ns Sy metadata
1452Controls what is cached in the primary cache
1453.Pq ARC .
1454If this property is set to
1455.Sy all ,
1456then both user data and metadata is cached.
1457If this property is set to
1458.Sy none ,
1459then neither user data nor metadata is cached.
1460If this property is set to
1461.Sy metadata ,
1462then only metadata is cached.
1463The default value is
1464.Sy all .
1465.It Sy quota Ns = Ns Em size Ns | Ns Sy none
1466Limits the amount of space a dataset and its descendents can consume.
1467This property enforces a hard limit on the amount of space used.
1468This includes all space consumed by descendents, including file systems and
1469snapshots.
1470Setting a quota on a descendent of a dataset that already has a quota does not
1471override the ancestor's quota, but rather imposes an additional limit.
1472.Pp
1473Quotas cannot be set on volumes, as the
1474.Sy volsize
1475property acts as an implicit quota.
1476.It Sy snapshot_limit Ns = Ns Em count Ns | Ns Sy none
1477Limits the number of snapshots that can be created on a dataset and its
1478descendents.
1479Setting a
1480.Sy snapshot_limit
1481on a descendent of a dataset that already has a
1482.Sy snapshot_limit
1483does not override the ancestor's
1484.Sy snapshot_limit ,
1485but rather imposes an additional limit.
1486The limit is not enforced if the user is allowed to change the limit.
1487For example, this means that recursive snapshots taken from the global zone are
1488counted against each delegated dataset within a zone.
1489This feature must be enabled to be used
1490.Po see
1491.Xr zpool-features 7
1492.Pc .
1493.It Sy userquota@ Ns Em user Ns = Ns Em size Ns | Ns Sy none
1494Limits the amount of space consumed by the specified user.
1495User space consumption is identified by the
1496.Sy userspace@ Ns Em user
1497property.
1498.Pp
1499Enforcement of user quotas may be delayed by several seconds.
1500This delay means that a user might exceed their quota before the system notices
1501that they are over quota and begins to refuse additional writes with the
1502.Er EDQUOT
1503error message.
1504See the
1505.Nm zfs Cm userspace
1506subcommand for more information.
1507.Pp
1508Unprivileged users can only access their own groups' space usage.
1509The root user, or a user who has been granted the
1510.Sy userquota
1511privilege with
1512.Nm zfs Cm allow ,
1513can get and set everyone's quota.
1514.Pp
1515This property is not available on volumes, on file systems before version 4, or
1516on pools before version 15.
1517The
1518.Sy userquota@ Ns Em ...
1519properties are not displayed by
1520.Nm zfs Cm get Sy all .
1521The user's name must be appended after the
1522.Sy @
1523symbol, using one of the following forms:
1524.Bl -bullet
1525.It
1526.Em POSIX name
1527.Po for example,
1528.Sy joe
1529.Pc
1530.It
1531.Em POSIX numeric ID
1532.Po for example,
1533.Sy 789
1534.Pc
1535.It
1536.Em SID name
1537.Po for example,
1538.Sy joe.smith@mydomain
1539.Pc
1540.It
1541.Em SID numeric ID
1542.Po for example,
1543.Sy S-1-123-456-789
1544.Pc
1545.El
1546.It Sy userobjquota@ Ns Em user Ns = Ns Em size Ns | Ns Sy none
1547The
1548.Sy userobjquota
1549is similar to
1550.Sy userquota
1551but it limits the number of objects a user can create.
1552Please refer to
1553.Sy userobjused
1554for more information about how objects are counted.
1555.It Sy groupquota@ Ns Em group Ns = Ns Em size Ns | Ns Sy none
1556Limits the amount of space consumed by the specified group.
1557Group space consumption is identified by the
1558.Sy groupused@ Ns Em group
1559property.
1560.Pp
1561Unprivileged users can access only their own groups' space usage.
1562The root user, or a user who has been granted the
1563.Sy groupquota
1564privilege with
1565.Nm zfs Cm allow ,
1566can get and set all groups' quotas.
1567.It Sy groupobjquota@ Ns Em group Ns = Ns Em size Ns | Ns Sy none
1568The
1569.Sy groupobjquota
1570is similar to
1571.Sy groupquota
1572but it limits the number of objects a group can consume.
1573Please refer to
1574.Sy userobjused
1575for more information about how objects are counted.
1576.It Sy projectquota@ Ns Em project Ns = Ns Em size Ns | Ns Sy none
1577Limits the amount of space consumed by the specified project.
1578Project space consumption is identified by the
1579.Sy projectused@ Ns Em project
1580property.
1581Please refer to
1582.Sy projectused
1583for more information about how project is identified and set or changed.
1584.Pp
1585The root user, or a user who has been granted the
1586.Sy projectquota
1587privilege with
1588.Nm zfs allow ,
1589can access all projects' quotas.
1590.It Sy projectobjquota@ Ns Em project Ns = Ns Em size Ns | Ns Sy none
1591The
1592.Sy projectobjquota
1593is similar to
1594.Sy projectquota
1595but it limits the number of objects a project can consume.
1596Please refer to
1597.Sy userobjused
1598for more information about how objects are counted.
1599.It Sy readonly Ns = Ns Sy on Ns | Ns Sy off
1600Controls whether this dataset can be modified.
1601The default value is
1602.Sy off .
1603.Pp
1604This property can also be referred to by its shortened column name,
1605.Sy rdonly .
1606.It Sy recordsize Ns = Ns Em size
1607Specifies a suggested block size for files in the file system.
1608This property is designed solely for use with database workloads that access
1609files in fixed-size records.
1610ZFS automatically tunes block sizes according to internal algorithms optimized
1611for typical access patterns.
1612.Pp
1613For databases that create very large files but access them in small random
1614chunks, these algorithms may be suboptimal.
1615Specifying a
1616.Sy recordsize
1617greater than or equal to the record size of the database can result in
1618significant performance gains.
1619Use of this property for general purpose file systems is strongly discouraged,
1620and may adversely affect performance.
1621.Pp
1622The size specified must be a power of two greater than or equal to 512 and less
1623than or equal to 128 Kbytes.
1624If the
1625.Sy large_blocks
1626feature is enabled on the pool, the size may be up to 1 Mbyte.
1627See
1628.Xr zpool-features 7
1629for details on ZFS feature flags.
1630.Pp
1631Changing the file system's
1632.Sy recordsize
1633affects only files created afterward; existing files are unaffected.
1634.Pp
1635This property can also be referred to by its shortened column name,
1636.Sy recsize .
1637.It Sy redundant_metadata Ns = Ns Sy all Ns | Ns Sy most
1638Controls what types of metadata are stored redundantly.
1639ZFS stores an extra copy of metadata, so that if a single block is corrupted,
1640the amount of user data lost is limited.
1641This extra copy is in addition to any redundancy provided at the pool level
1642.Pq e.g. by mirroring or RAID-Z ,
1643and is in addition to an extra copy specified by the
1644.Sy copies
1645property
1646.Pq up to a total of 3 copies .
1647For example if the pool is mirrored,
1648.Sy copies Ns = Ns 2 ,
1649and
1650.Sy redundant_metadata Ns = Ns Sy most ,
1651then ZFS stores 6 copies of most metadata, and 4 copies of data and some
1652metadata.
1653.Pp
1654When set to
1655.Sy all ,
1656ZFS stores an extra copy of all metadata.
1657If a single on-disk block is corrupt, at worst a single block of user data
1658.Po which is
1659.Sy recordsize
1660bytes long
1661.Pc
1662can be lost.
1663.Pp
1664When set to
1665.Sy most ,
1666ZFS stores an extra copy of most types of metadata.
1667This can improve performance of random writes, because less metadata must be
1668written.
1669In practice, at worst about 100 blocks
1670.Po of
1671.Sy recordsize
1672bytes each
1673.Pc
1674of user data can be lost if a single on-disk block is corrupt.
1675The exact behavior of which metadata blocks are stored redundantly may change in
1676future releases.
1677.Pp
1678The default value is
1679.Sy all .
1680.It Sy refquota Ns = Ns Em size Ns | Ns Sy none
1681Limits the amount of space a dataset can consume.
1682This property enforces a hard limit on the amount of space used.
1683This hard limit does not include space used by descendents, including file
1684systems and snapshots.
1685.It Sy refreservation Ns = Ns Em size Ns | Ns Sy none Ns | Ns Sy auto
1686The minimum amount of space guaranteed to a dataset, not including its
1687descendents.
1688When the amount of space used is below this value, the dataset is treated as if
1689it were taking up the amount of space specified by
1690.Sy refreservation .
1691The
1692.Sy refreservation
1693reservation is accounted for in the parent datasets' space used, and counts
1694against the parent datasets' quotas and reservations.
1695.Pp
1696If
1697.Sy refreservation
1698is set, a snapshot is only allowed if there is enough free pool space outside of
1699this reservation to accommodate the current number of
1700.Qq referenced
1701bytes in the dataset.
1702.Pp
1703If
1704.Sy refreservation
1705is set to
1706.Sy auto ,
1707a volume is thick provisioned
1708.Po or
1709.Qq not sparse
1710.Pc .
1711.Sy refreservation Ns = Ns Sy auto
1712is only supported on volumes.
1713See
1714.Sy volsize
1715in the
1716.Sx Native Properties
1717section for more information about sparse volumes.
1718.Pp
1719This property can also be referred to by its shortened column name,
1720.Sy refreserv .
1721.It Sy reservation Ns = Ns Em size Ns | Ns Sy none
1722The minimum amount of space guaranteed to a dataset and its descendants.
1723When the amount of space used is below this value, the dataset is treated as if
1724it were taking up the amount of space specified by its reservation.
1725Reservations are accounted for in the parent datasets' space used, and count
1726against the parent datasets' quotas and reservations.
1727.Pp
1728This property can also be referred to by its shortened column name,
1729.Sy reserv .
1730.It Sy secondarycache Ns = Ns Sy all Ns | Ns Sy none Ns | Ns Sy metadata
1731Controls what is cached in the secondary cache
1732.Pq L2ARC .
1733If this property is set to
1734.Sy all ,
1735then both user data and metadata is cached.
1736If this property is set to
1737.Sy none ,
1738then neither user data nor metadata is cached.
1739If this property is set to
1740.Sy metadata ,
1741then only metadata is cached.
1742The default value is
1743.Sy all .
1744.It Sy setuid Ns = Ns Sy on Ns | Ns Sy off
1745Controls whether the setuid bit is respected for the file system.
1746The default value is
1747.Sy on .
1748.It Sy sharesmb Ns = Ns Sy on Ns | Ns Sy off Ns | Ns Em opts
1749Controls whether the file system is shared via SMB, and what options are to be
1750used.
1751A file system with the
1752.Sy sharesmb
1753property set to
1754.Sy off
1755is managed through traditional tools such as
1756.Xr sharemgr 8 .
1757Otherwise, the file system is automatically shared and unshared with the
1758.Nm zfs Cm share
1759and
1760.Nm zfs Cm unshare
1761commands.
1762If the property is set to
1763.Sy on ,
1764the
1765.Xr sharemgr 8
1766command is invoked with no options.
1767Otherwise, the
1768.Xr sharemgr 8
1769command is invoked with options equivalent to the contents of this property.
1770.Pp
1771Because SMB shares requires a resource name, a unique resource name is
1772constructed from the dataset name.
1773The constructed name is a copy of the dataset name except that the characters in
1774the dataset name, which would be invalid in the resource name, are replaced with
1775underscore
1776.Pq Sy _
1777characters.
1778A pseudo property
1779.Qq name
1780is also supported that allows you to replace the data set name with a specified
1781name.
1782The specified name is then used to replace the prefix dataset in the case of
1783inheritance.
1784For example, if the dataset
1785.Em data/home/john
1786is set to
1787.Sy name Ns = Ns Sy john ,
1788then
1789.Em data/home/john
1790has a resource name of
1791.Sy john .
1792If a child dataset
1793.Em data/home/john/backups
1794is shared, it has a resource name of
1795.Sy john_backups .
1796.Pp
1797When SMB shares are created, the SMB share name appears as an entry in the
1798.Pa .zfs/shares
1799directory.
1800You can use the
1801.Nm ls
1802or
1803.Nm chmod
1804command to display the share-level ACLs on the entries in this directory.
1805.Pp
1806When the
1807.Sy sharesmb
1808property is changed for a dataset, the dataset and any children inheriting the
1809property are re-shared with the new options, only if the property was previously
1810set to
1811.Sy off ,
1812or if they were shared before the property was changed.
1813If the new property is set to
1814.Sy off ,
1815the file systems are unshared.
1816.It Sy sharenfs Ns = Ns Sy on Ns | Ns Sy off Ns | Ns Em opts
1817Controls whether the file system is shared via NFS, and what options are to be
1818used.
1819A file system with a
1820.Sy sharenfs
1821property of
1822.Sy off
1823is managed through traditional tools such as
1824.Xr share 8 ,
1825.Xr unshare 8 ,
1826and
1827.Xr dfstab 5 .
1828Otherwise, the file system is automatically shared and unshared with the
1829.Nm zfs Cm share
1830and
1831.Nm zfs Cm unshare
1832commands.
1833If the property is set to
1834.Sy on ,
1835.Xr share 8
1836command is invoked with no options.
1837Otherwise, the
1838.Xr share 8
1839command is invoked with options equivalent to the contents of this property.
1840.Pp
1841When the
1842.Sy sharenfs
1843property is changed for a dataset, the dataset and any children inheriting the
1844property are re-shared with the new options, only if the property was previously
1845.Sy off ,
1846or if they were shared before the property was changed.
1847If the new property is
1848.Sy off ,
1849the file systems are unshared.
1850.It Sy logbias Ns = Ns Sy latency Ns | Ns Sy throughput
1851Provide a hint to ZFS about handling of synchronous requests in this dataset.
1852If
1853.Sy logbias
1854is set to
1855.Sy latency
1856.Pq the default ,
1857ZFS will use pool log devices
1858.Pq if configured
1859to handle the requests at low latency.
1860If
1861.Sy logbias
1862is set to
1863.Sy throughput ,
1864ZFS will not use configured pool log devices.
1865ZFS will instead optimize synchronous operations for global pool throughput and
1866efficient use of resources.
1867.It Sy snapdir Ns = Ns Sy hidden Ns | Ns Sy visible
1868Controls whether the
1869.Pa .zfs
1870directory is hidden or visible in the root of the file system as discussed in
1871the
1872.Sx Snapshots
1873section.
1874The default value is
1875.Sy hidden .
1876.It Sy sync Ns = Ns Sy standard Ns | Ns Sy always Ns | Ns Sy disabled
1877Controls the behavior of synchronous requests
1878.Pq e.g. fsync, O_DSYNC .
1879.Sy standard
1880is the
1881POSIX
1882specified behavior of ensuring all synchronous requests are written to stable
1883storage and all devices are flushed to ensure data is not cached by device
1884controllers
1885.Pq this is the default .
1886.Sy always
1887causes every file system transaction to be written and flushed before its
1888system call returns.
1889This has a large performance penalty.
1890.Sy disabled
1891disables synchronous requests.
1892File system transactions are only committed to stable storage periodically.
1893This option will give the highest performance.
1894However, it is very dangerous as ZFS would be ignoring the synchronous
1895transaction demands of applications such as databases or NFS.
1896Administrators should only use this option when the risks are understood.
1897.It Sy version Ns = Ns Em N Ns | Ns Sy current
1898The on-disk version of this file system, which is independent of the pool
1899version.
1900This property can only be set to later supported versions.
1901See the
1902.Nm zfs Cm upgrade
1903command.
1904.It Sy volsize Ns = Ns Em size
1905For volumes, specifies the logical size of the volume.
1906By default, creating a volume establishes a reservation of equal size.
1907For storage pools with a version number of 9 or higher, a
1908.Sy refreservation
1909is set instead.
1910Any changes to
1911.Sy volsize
1912are reflected in an equivalent change to the reservation
1913.Po or
1914.Sy refreservation
1915.Pc .
1916The
1917.Sy volsize
1918can only be set to a multiple of
1919.Sy volblocksize ,
1920and cannot be zero.
1921.Pp
1922The reservation is kept equal to the volume's logical size to prevent unexpected
1923behavior for consumers.
1924Without the reservation, the volume could run out of space, resulting in
1925undefined behavior or data corruption, depending on how the volume is used.
1926These effects can also occur when the volume size is changed while it is in use
1927.Pq particularly when shrinking the size .
1928Extreme care should be used when adjusting the volume size.
1929.Pp
1930Though not recommended, a
1931.Qq sparse volume
1932.Po also known as
1933.Qq thin provisioned
1934.Pc
1935can be created by specifying the
1936.Fl s
1937option to the
1938.Nm zfs Cm create Fl V
1939command, or by changing the value of the
1940.Sy refreservation
1941property
1942.Po or
1943.Sy reservation
1944property on pool version 8 or earlier
1945.Pc
1946after the volume has been created.
1947A
1948.Qq sparse volume
1949is a volume where the value of
1950.Sy refreservation
1951is less than the size of the volume plus the space required to store its
1952metadata.
1953Consequently, writes to a sparse volume can fail with
1954.Er ENOSPC
1955when the pool is low on space.
1956For a sparse volume, changes to
1957.Sy volsize
1958are not reflected in the
1959.Sy refreservation .
1960A volume that is not sparse is said to be
1961.Qq thick provisioned .
1962A sparse volume can become thick provisioned by setting
1963.Sy refreservation
1964to
1965.Sy auto .
1966.It Sy vscan Ns = Ns Sy on Ns | Ns Sy off
1967Controls whether regular files should be scanned for viruses when a file is
1968opened and closed.
1969In addition to enabling this property, the virus scan service must also be
1970enabled for virus scanning to occur.
1971The default value is
1972.Sy off .
1973.It Sy xattr Ns = Ns Sy on Ns | Ns Sy off
1974Controls whether extended attributes are enabled for this file system.
1975The default value is
1976.Sy on .
1977.It Sy zoned Ns = Ns Sy on Ns | Ns Sy off
1978Controls whether the dataset is managed from a non-global zone.
1979See the
1980.Sx Zones
1981section for more information.
1982The default value is
1983.Sy off .
1984.El
1985.Pp
1986The following three properties cannot be changed after the file system is
1987created, and therefore, should be set when the file system is created.
1988If the properties are not set with the
1989.Nm zfs Cm create
1990or
1991.Nm zpool Cm create
1992commands, these properties are inherited from the parent dataset.
1993If the parent dataset lacks these properties due to having been created prior to
1994these features being supported, the new file system will have the default values
1995for these properties.
1996.Bl -tag -width ""
1997.It Xo
1998.Sy casesensitivity Ns = Ns Sy sensitive Ns | Ns
1999.Sy insensitive Ns | Ns Sy mixed
2000.Xc
2001Indicates whether the file name matching algorithm used by the file system
2002should be case-sensitive, case-insensitive, or allow a combination of both
2003styles of matching.
2004The default value for the
2005.Sy casesensitivity
2006property is
2007.Sy sensitive .
2008Traditionally,
2009.Ux
2010and
2011POSIX
2012file systems have case-sensitive file names.
2013.Pp
2014The
2015.Sy mixed
2016value for the
2017.Sy casesensitivity
2018property indicates that the file system can support requests for both
2019case-sensitive and case-insensitive matching behavior.
2020Currently, case-insensitive matching behavior on a file system that supports
2021mixed behavior is limited to the SMB server product.
2022For more information about the
2023.Sy mixed
2024value behavior, see the "ZFS Administration Guide".
2025.It Xo
2026.Sy normalization Ns = Ns Sy none Ns | Ns Sy formC Ns | Ns
2027.Sy formD Ns | Ns Sy formKC Ns | Ns Sy formKD
2028.Xc
2029Indicates whether the file system should perform a
2030.Sy unicode
2031normalization of file names whenever two file names are compared, and which
2032normalization algorithm should be used.
2033File names are always stored unmodified, names are normalized as part of any
2034comparison process.
2035If this property is set to a legal value other than
2036.Sy none ,
2037and the
2038.Sy utf8only
2039property was left unspecified, the
2040.Sy utf8only
2041property is automatically set to
2042.Sy on .
2043The default value of the
2044.Sy normalization
2045property is
2046.Sy none .
2047This property cannot be changed after the file system is created.
2048.It Sy utf8only Ns = Ns Sy on Ns | Ns Sy off
2049Indicates whether the file system should reject file names that include
2050characters that are not present in the
2051.Sy UTF-8
2052character code set.
2053If this property is explicitly set to
2054.Sy off ,
2055the normalization property must either not be explicitly set or be set to
2056.Sy none .
2057The default value for the
2058.Sy utf8only
2059property is
2060.Sy off .
2061This property cannot be changed after the file system is created.
2062.El
2063.Pp
2064The
2065.Sy casesensitivity ,
2066.Sy normalization ,
2067and
2068.Sy utf8only
2069properties are also new permissions that can be assigned to non-privileged users
2070by using the ZFS delegated administration feature.
2071.Ss "Temporary Mount Point Properties"
2072When a file system is mounted, either through
2073.Xr mount 8
2074for legacy mounts or the
2075.Nm zfs Cm mount
2076command for normal file systems, its mount options are set according to its
2077properties.
2078The correlation between properties and mount options is as follows:
2079.Bd -literal
2080    PROPERTY                MOUNT OPTION
2081    devices                 devices/nodevices
2082    exec                    exec/noexec
2083    readonly                ro/rw
2084    setuid                  setuid/nosetuid
2085    xattr                   xattr/noxattr
2086.Ed
2087.Pp
2088In addition, these options can be set on a per-mount basis using the
2089.Fl o
2090option, without affecting the property that is stored on disk.
2091The values specified on the command line override the values stored in the
2092dataset.
2093The
2094.Sy nosuid
2095option is an alias for
2096.Sy nodevices Ns \&, Ns Sy nosetuid .
2097These properties are reported as
2098.Qq temporary
2099by the
2100.Nm zfs Cm get
2101command.
2102If the properties are changed while the dataset is mounted, the new setting
2103overrides any temporary settings.
2104.Ss "User Properties"
2105In addition to the standard native properties, ZFS supports arbitrary user
2106properties.
2107User properties have no effect on ZFS behavior, but applications or
2108administrators can use them to annotate datasets
2109.Pq file systems, volumes, and snapshots .
2110.Pp
2111User property names must contain a colon
2112.Pq Qq Sy \&:
2113character to distinguish them from native properties.
2114They may contain lowercase letters, numbers, and the following punctuation
2115characters: colon
2116.Pq Qq Sy \&: ,
2117dash
2118.Pq Qq Sy - ,
2119period
2120.Pq Qq Sy \&. ,
2121and underscore
2122.Pq Qq Sy _ .
2123The expected convention is that the property name is divided into two portions
2124such as
2125.Em module Ns \&: Ns Em property ,
2126but this namespace is not enforced by ZFS.
2127User property names can be at most 256 characters, and cannot begin with a dash
2128.Pq Qq Sy - .
2129.Pp
2130When making programmatic use of user properties, it is strongly suggested to use
2131a reversed
2132.Sy DNS
2133domain name for the
2134.Em module
2135component of property names to reduce the chance that two
2136independently-developed packages use the same property name for different
2137purposes.
2138.Pp
2139The values of user properties are arbitrary strings, are always inherited, and
2140are never validated.
2141All of the commands that operate on properties
2142.Po Nm zfs Cm list ,
2143.Nm zfs Cm get ,
2144.Nm zfs Cm set ,
2145and so forth
2146.Pc
2147can be used to manipulate both native properties and user properties.
2148Use the
2149.Nm zfs Cm inherit
2150command to clear a user property.
2151If the property is not defined in any parent dataset, it is removed entirely.
2152Property values are limited to 8192 bytes.
2153.Ss ZFS Volumes as Swap or Dump Devices
2154During an initial installation a swap device and dump device are created on ZFS
2155volumes in the ZFS root pool.
2156By default, the swap area size is based on 1/2 the size of physical memory up to
21572 Gbytes.
2158The size of the dump device depends on the kernel's requirements at installation
2159time.
2160Separate ZFS volumes must be used for the swap area and dump devices.
2161Do not swap to a file on a ZFS file system.
2162A ZFS swap file configuration is not supported.
2163.Pp
2164If you need to change your swap area or dump device after the system is
2165installed or upgraded, use the
2166.Xr swap 8
2167and
2168.Xr dumpadm 8
2169commands.
2170.Ss "Encryption"
2171Enabling the
2172.Sy encryption
2173feature allows for the creation of encrypted filesystems and volumes.
2174ZFS will encrypt all user data including file and zvol data, file attributes,
2175ACLs, permission bits, directory listings, FUID mappings, and userused/groupused
2176data.
2177ZFS
2178will not encrypt metadata related to the pool structure, including dataset
2179names, dataset hierarchy, file size, file holes, and dedup tables.
2180Key rotation is managed internally by the ZFS kernel module and changing the
2181user's key does not require re-encrypting the entire dataset.
2182Datasets can be scrubbed, resilvered, renamed, and deleted without the
2183encryption keys being loaded
2184.Po see the
2185.Nm Cm load-key
2186subcommand for more info on key loading
2187.Pc .
2188.Pp
2189Creating an encrypted dataset requires specifying the
2190.Sy encryption
2191and
2192.Sy keyformat
2193properties at creation time, along with an optional
2194.Sy keylocation
2195and
2196.Sy pbkdf2iters .
2197After entering an encryption key, the created
2198dataset will become an encryption root.
2199Any descendant datasets will inherit their encryption key from the encryption
2200root by default, meaning that loading, unloading, or changing the key for the
2201encryption root will implicitly do the same for all inheriting datasets.
2202If this inheritance is not desired, simply supply a
2203.Sy keyformat
2204when creating the child dataset or use
2205.Nm Cm change-key
2206to break an existing relationship, creating a new encryption root on the child.
2207Note that the child's
2208.Sy keyformat
2209may match that of the parent while still creating a new encryption root, and
2210that changing the
2211.Sy encryption
2212property alone does not create a new encryption root; this would simply use a
2213different cipher suite with the same key as its encryption root.
2214The one exception is that clones will always use their origin's encryption key.
2215As a result of this exception, some encryption-related properties (namely
2216.Sy keystatus ,
2217.Sy keyformat ,
2218.Sy keylocation ,
2219and
2220.Sy pbkdf2iters )
2221do not inherit like other ZFS properties and instead use the value determined
2222by their encryption root.
2223Encryption root inheritance can be tracked via the read-only
2224.Sy encryptionroot
2225property.
2226.Pp
2227Encryption changes the behavior of a few ZFS operations.
2228Encryption is applied after compression so compression ratios are preserved.
2229Normally checksums in ZFS are 256 bits long, but for encrypted data the checksum
2230is 128 bits of the user-chosen checksum and 128 bits of MAC from the encryption
2231suite, which provides additional protection against maliciously altered data.
2232Deduplication is still possible with encryption enabled but for security,
2233datasets will only dedup against themselves, their snapshots, and their clones.
2234.Pp
2235There are a few limitations on encrypted datasets.
2236Encrypted data cannot be embedded via the
2237.Sy embedded_data
2238feature.
2239Encrypted datasets may not have
2240.Sy copies Ns = Ns Sy 3
2241since the implementation stores some encryption metadata where the third copy
2242would normally be.
2243Since compression is applied before encryption datasets may be vulnerable to a
2244CRIME-like attack if applications accessing the data allow for it.
2245Deduplication with encryption will leak information about which blocks are
2246equivalent in a dataset and will incur an extra CPU cost per block written.
2247.Sh SUBCOMMANDS
2248All subcommands that modify state are logged persistently to the pool in their
2249original form.
2250.Bl -tag -width ""
2251.It Nm Fl \&?
2252Displays a help message.
2253.It Xo
2254.Nm
2255.Cm create
2256.Op Fl Pnpv
2257.Oo Fl o Ar property Ns = Ns Ar value Oc Ns ...
2258.Ar filesystem
2259.Xc
2260Creates a new ZFS file system.
2261The file system is automatically mounted according to the
2262.Sy mountpoint
2263property inherited from the parent.
2264.Bl -tag -width "-o"
2265.It Fl o Ar property Ns = Ns Ar value
2266Sets the specified property as if the command
2267.Nm zfs Cm set Ar property Ns = Ns Ar value
2268was invoked at the same time the dataset was created.
2269Any editable ZFS property can also be set at creation time.
2270Multiple
2271.Fl o
2272options can be specified.
2273An error results if the same property is specified in multiple
2274.Fl o
2275options.
2276.It Fl p
2277Creates all the non-existing parent datasets.
2278Datasets created in this manner are automatically mounted according to the
2279.Sy mountpoint
2280property inherited from their parent.
2281Any property specified on the command line using the
2282.Fl o
2283option is ignored.
2284If the target filesystem already exists, the operation completes successfully.
2285.It Fl n
2286Do a dry-run
2287.Pq Qq No-op
2288creation.
2289No datasets will be created.
2290This is useful in conjunction with the
2291.Fl v
2292or
2293.Fl P
2294flags to validate properties that are passed via
2295.Fl o
2296options and those implied by other options.
2297The actual dataset creation can still fail due to insufficient privileges or
2298available capacity.
2299.It Fl P
2300Print machine-parsable verbose information about the created dataset.
2301Each line of output contains a key and one or two values, all separated by tabs.
2302The
2303.Sy create_ancestors
2304and
2305.Sy create
2306keys have
2307.Em filesystem
2308as their only value.
2309The
2310.Sy create_ancestors
2311key only appears if the
2312.Fl p
2313option is used.
2314The
2315.Sy property
2316key has two values, a property's name and that property's value.
2317The
2318.Sy property
2319key may appear zero or more times, once for each property that will be set local
2320to
2321.Em filesystem
2322due to the use of the
2323.Fl o
2324option.
2325.It Fl v
2326Print verbose information about the created dataset.
2327.El
2328.It Xo
2329.Nm
2330.Cm create
2331.Op Fl ps
2332.Op Fl b Ar blocksize
2333.Oo Fl o Ar property Ns = Ns Ar value Oc Ns ...
2334.Fl V Ar size Ar volume
2335.Xc
2336Creates a volume of the given size.
2337The volume is exported as a block device in
2338.Pa /dev/zvol/{dsk,rdsk}/path ,
2339where
2340.Em path
2341is the name of the volume in the ZFS namespace.
2342The size represents the logical size as exported by the device.
2343By default, a reservation of equal size is created.
2344.Pp
2345.Ar size
2346is automatically rounded up to the nearest 128 Kbytes to ensure that the volume
2347has an integral number of blocks regardless of
2348.Sy blocksize .
2349.Bl -tag -width "-b"
2350.It Fl b Ar blocksize
2351Equivalent to
2352.Fl o Sy volblocksize Ns = Ns Ar blocksize .
2353If this option is specified in conjunction with
2354.Fl o Sy volblocksize ,
2355the resulting behavior is undefined.
2356.It Fl o Ar property Ns = Ns Ar value
2357Sets the specified property as if the
2358.Nm zfs Cm set Ar property Ns = Ns Ar value
2359command was invoked at the same time the dataset was created.
2360Any editable ZFS property can also be set at creation time.
2361Multiple
2362.Fl o
2363options can be specified.
2364An error results if the same property is specified in multiple
2365.Fl o
2366options.
2367.It Fl p
2368Creates all the non-existing parent datasets.
2369Datasets created in this manner are automatically mounted according to the
2370.Sy mountpoint
2371property inherited from their parent.
2372Any property specified on the command line using the
2373.Fl o
2374option is ignored.
2375If the target filesystem already exists, the operation completes successfully.
2376.It Fl s
2377Creates a sparse volume with no reservation.
2378See
2379.Sy volsize
2380in the
2381.Sx Native Properties
2382section for more information about sparse volumes.
2383.It Fl n
2384Do a dry-run
2385.Pq Qq No-op
2386creation.
2387No datasets will be created.
2388This is useful in conjunction with the
2389.Fl v
2390or
2391.Fl P
2392flags to validate properties that are passed via
2393.Fl o
2394options and those implied by other options.
2395The actual dataset creation can still fail due to insufficient privileges or
2396available capacity.
2397.It Fl P
2398Print machine-parsable verbose information about the created dataset.
2399Each line of output contains a key and one or two values, all separated by tabs.
2400The
2401.Sy create_ancestors
2402and
2403.Sy create
2404keys have
2405.Em volume
2406as their only value.
2407The
2408.Sy create_ancestors
2409key only appears if the
2410.Fl p
2411option is used.
2412The
2413.Sy property
2414key has two values, a property's name and that property's value.
2415The
2416.Sy property
2417key may appear zero or more times, once for each property that will be set local
2418to
2419.Em volume
2420due to the use of the
2421.Fl b
2422or
2423.Fl o
2424options, as well as
2425.Sy refreservation
2426if the volume is not sparse.
2427.It Fl v
2428Print verbose information about the created dataset.
2429.El
2430.It Xo
2431.Nm
2432.Cm destroy
2433.Op Fl Rfnprv
2434.Ar filesystem Ns | Ns Ar volume
2435.Xc
2436Destroys the given dataset.
2437By default, the command unshares any file systems that are currently shared,
2438unmounts any file systems that are currently mounted, and refuses to destroy a
2439dataset that has active dependents
2440.Pq children or clones .
2441.Bl -tag -width "-R"
2442.It Fl R
2443Recursively destroy all dependents, including cloned file systems outside the
2444target hierarchy.
2445.It Fl f
2446Force an unmount of any file systems using the
2447.Nm unmount Fl f
2448command.
2449This option has no effect on non-file systems or unmounted file systems.
2450.It Fl n
2451Do a dry-run
2452.Pq Qq No-op
2453deletion.
2454No data will be deleted.
2455This is useful in conjunction with the
2456.Fl v
2457or
2458.Fl p
2459flags to determine what data would be deleted.
2460.It Fl p
2461Print machine-parsable verbose information about the deleted data.
2462.It Fl r
2463Recursively destroy all children.
2464.It Fl v
2465Print verbose information about the deleted data.
2466.El
2467.Pp
2468Extreme care should be taken when applying either the
2469.Fl r
2470or the
2471.Fl R
2472options, as they can destroy large portions of a pool and cause unexpected
2473behavior for mounted file systems in use.
2474.It Xo
2475.Nm
2476.Cm destroy
2477.Op Fl Rdnprv
2478.Ar filesystem Ns | Ns Ar volume Ns @ Ns Ar snap Ns
2479.Oo % Ns Ar snap Ns Oo , Ns Ar snap Ns Oo % Ns Ar snap Oc Oc Oc Ns ...
2480.Xc
2481The given snapshots are destroyed immediately if and only if the
2482.Nm zfs Cm destroy
2483command without the
2484.Fl d
2485option would have destroyed it.
2486Such immediate destruction would occur, for example, if the snapshot had no
2487clones and the user-initiated reference count were zero.
2488.Pp
2489If a snapshot does not qualify for immediate destruction, it is marked for
2490deferred deletion.
2491In this state, it exists as a usable, visible snapshot until both of the
2492preconditions listed above are met, at which point it is destroyed.
2493.Pp
2494An inclusive range of snapshots may be specified by separating the first and
2495last snapshots with a percent sign.
2496The first and/or last snapshots may be left blank, in which case the
2497filesystem's oldest or newest snapshot will be implied.
2498.Pp
2499Multiple snapshots
2500.Pq or ranges of snapshots
2501of the same filesystem or volume may be specified in a comma-separated list of
2502snapshots.
2503Only the snapshot's short name
2504.Po the part after the
2505.Sy @
2506.Pc
2507should be specified when using a range or comma-separated list to identify
2508multiple snapshots.
2509.Bl -tag -width "-R"
2510.It Fl R
2511Recursively destroy all clones of these snapshots, including the clones,
2512snapshots, and children.
2513If this flag is specified, the
2514.Fl d
2515flag will have no effect.
2516.It Fl d
2517Defer snapshot deletion.
2518.It Fl n
2519Do a dry-run
2520.Pq Qq No-op
2521deletion.
2522No data will be deleted.
2523This is useful in conjunction with the
2524.Fl p
2525or
2526.Fl v
2527flags to determine what data would be deleted.
2528.It Fl p
2529Print machine-parsable verbose information about the deleted data.
2530.It Fl r
2531Destroy
2532.Pq or mark for deferred deletion
2533all snapshots with this name in descendent file systems.
2534.It Fl v
2535Print verbose information about the deleted data.
2536.Pp
2537Extreme care should be taken when applying either the
2538.Fl r
2539or the
2540.Fl R
2541options, as they can destroy large portions of a pool and cause unexpected
2542behavior for mounted file systems in use.
2543.El
2544.It Xo
2545.Nm
2546.Cm destroy
2547.Ar filesystem Ns | Ns Ar volume Ns # Ns Ar bookmark
2548.Xc
2549The given bookmark is destroyed.
2550.It Xo
2551.Nm
2552.Cm snapshot
2553.Op Fl r
2554.Oo Fl o Ar property Ns = Ns value Oc Ns ...
2555.Ar filesystem Ns @ Ns Ar snapname Ns | Ns Ar volume Ns @ Ns Ar snapname Ns ...
2556.Xc
2557Creates snapshots with the given names.
2558All previous modifications by successful system calls to the file system are
2559part of the snapshots.
2560Snapshots are taken atomically, so that all snapshots correspond to the same
2561moment in time.
2562See the
2563.Sx Snapshots
2564section for details.
2565.Bl -tag -width "-o"
2566.It Fl o Ar property Ns = Ns Ar value
2567Sets the specified property; see
2568.Nm zfs Cm create
2569for details.
2570.It Fl r
2571Recursively create snapshots of all descendent datasets
2572.El
2573.It Xo
2574.Nm
2575.Cm rollback
2576.Op Fl Rfr
2577.Ar snapshot
2578.Xc
2579Roll back the given dataset to a previous snapshot.
2580When a dataset is rolled back, all data that has changed since the snapshot is
2581discarded, and the dataset reverts to the state at the time of the snapshot.
2582By default, the command refuses to roll back to a snapshot other than the most
2583recent one.
2584In order to do so, all intermediate snapshots and bookmarks must be destroyed by
2585specifying the
2586.Fl r
2587option.
2588.Pp
2589The
2590.Fl rR
2591options do not recursively destroy the child snapshots of a recursive snapshot.
2592Only direct snapshots of the specified filesystem are destroyed by either of
2593these options.
2594To completely roll back a recursive snapshot, you must rollback the individual
2595child snapshots.
2596.Bl -tag -width "-R"
2597.It Fl R
2598Destroy any more recent snapshots and bookmarks, as well as any clones of those
2599snapshots.
2600.It Fl f
2601Used with the
2602.Fl R
2603option to force an unmount of any clone file systems that are to be destroyed.
2604.It Fl r
2605Destroy any snapshots and bookmarks more recent than the one specified.
2606.El
2607.It Xo
2608.Nm
2609.Cm clone
2610.Op Fl p
2611.Oo Fl o Ar property Ns = Ns Ar value Oc Ns ...
2612.Ar snapshot Ar filesystem Ns | Ns Ar volume
2613.Xc
2614Creates a clone of the given snapshot.
2615See the
2616.Sx Clones
2617section for details.
2618The target dataset can be located anywhere in the ZFS hierarchy, and is created
2619as the same type as the original.
2620.Bl -tag -width "-o"
2621.It Fl o Ar property Ns = Ns Ar value
2622Sets the specified property; see
2623.Nm zfs Cm create
2624for details.
2625.It Fl p
2626Creates all the non-existing parent datasets.
2627Datasets created in this manner are automatically mounted according to the
2628.Sy mountpoint
2629property inherited from their parent.
2630If the target filesystem or volume already exists, the operation completes
2631successfully.
2632.El
2633.It Xo
2634.Nm
2635.Cm promote
2636.Ar clone-filesystem
2637.Xc
2638Promotes a clone file system to no longer be dependent on its
2639.Qq origin
2640snapshot.
2641This makes it possible to destroy the file system that the clone was created
2642from.
2643The clone parent-child dependency relationship is reversed, so that the origin
2644file system becomes a clone of the specified file system.
2645.Pp
2646The snapshot that was cloned, and any snapshots previous to this snapshot, are
2647now owned by the promoted clone.
2648The space they use moves from the origin file system to the promoted clone, so
2649enough space must be available to accommodate these snapshots.
2650No new space is consumed by this operation, but the space accounting is
2651adjusted.
2652The promoted clone must not have any conflicting snapshot names of its own.
2653The
2654.Cm rename
2655subcommand can be used to rename any conflicting snapshots.
2656.It Xo
2657.Nm
2658.Cm rename
2659.Op Fl f
2660.Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot
2661.Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot
2662.Xc
2663.It Xo
2664.Nm
2665.Cm rename
2666.Op Fl fp
2667.Ar filesystem Ns | Ns Ar volume
2668.Ar filesystem Ns | Ns Ar volume
2669.Xc
2670Renames the given dataset.
2671The new target can be located anywhere in the ZFS hierarchy, with the exception
2672of snapshots.
2673Snapshots can only be renamed within the parent file system or volume.
2674When renaming a snapshot, the parent file system of the snapshot does not need
2675to be specified as part of the second argument.
2676Renamed file systems can inherit new mount points, in which case they are
2677unmounted and remounted at the new mount point.
2678.Bl -tag -width "-a"
2679.It Fl f
2680Force unmount any filesystems that need to be unmounted in the process.
2681.It Fl p
2682Creates all the nonexistent parent datasets.
2683Datasets created in this manner are automatically mounted according to the
2684.Sy mountpoint
2685property inherited from their parent.
2686.El
2687.It Xo
2688.Nm
2689.Cm rename
2690.Fl r
2691.Ar snapshot Ar snapshot
2692.Xc
2693Recursively rename the snapshots of all descendent datasets.
2694Snapshots are the only dataset that can be renamed recursively.
2695.It Xo
2696.Nm
2697.Cm list
2698.Op Fl r Ns | Ns Fl d Ar depth
2699.Op Fl Hp
2700.Oo Fl o Ar property Ns Oo , Ns Ar property Oc Ns ... Oc
2701.Oo Fl s Ar property Oc Ns ...
2702.Oo Fl S Ar property Oc Ns ...
2703.Oo Fl t Ar type Ns Oo , Ns Ar type Oc Ns ... Oc
2704.Oo Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot Oc Ns ...
2705.Xc
2706Lists the property information for the given datasets in tabular form.
2707If specified, you can list property information by the absolute pathname or the
2708relative pathname.
2709By default, all file systems and volumes are displayed.
2710Snapshots are displayed if the
2711.Sy listsnaps
2712property is
2713.Sy on
2714.Po the default is
2715.Sy off
2716.Pc .
2717The following fields are displayed,
2718.Sy name Ns \&, Ns Sy used Ns \&, Ns Sy available Ns \&, Ns Sy referenced Ns \&, Ns
2719.Sy mountpoint .
2720.Bl -tag -width "-H"
2721.It Fl H
2722Used for scripting mode.
2723Do not print headers and separate fields by a single tab instead of arbitrary
2724white space.
2725.It Fl S Ar property
2726Same as the
2727.Fl s
2728option, but sorts by property in descending order.
2729.It Fl d Ar depth
2730Recursively display any children of the dataset, limiting the recursion to
2731.Ar depth .
2732A
2733.Ar depth
2734of
2735.Sy 1
2736will display only the dataset and its direct children.
2737.It Fl o Ar property
2738A comma-separated list of properties to display.
2739The property must be:
2740.Bl -bullet
2741.It
2742One of the properties described in the
2743.Sx Native Properties
2744section
2745.It
2746A user property
2747.It
2748The value
2749.Sy name
2750to display the dataset name
2751.It
2752The value
2753.Sy space
2754to display space usage properties on file systems and volumes.
2755This is a shortcut for specifying
2756.Fl o Sy name Ns \&, Ns Sy avail Ns \&, Ns Sy used Ns \&, Ns Sy usedsnap Ns \&, Ns
2757.Sy usedds Ns \&, Ns Sy usedrefreserv Ns \&, Ns Sy usedchild Fl t
2758.Sy filesystem Ns \&, Ns Sy volume
2759syntax.
2760.El
2761.It Fl p
2762Display numbers in parsable
2763.Pq exact
2764values.
2765.It Fl r
2766Recursively display any children of the dataset on the command line.
2767.It Fl s Ar property
2768A property for sorting the output by column in ascending order based on the
2769value of the property.
2770The property must be one of the properties described in the
2771.Sx Properties
2772section, or the special value
2773.Sy name
2774to sort by the dataset name.
2775Multiple properties can be specified at one time using multiple
2776.Fl s
2777property options.
2778Multiple
2779.Fl s
2780options are evaluated from left to right in decreasing order of importance.
2781The following is a list of sorting criteria:
2782.Bl -bullet
2783.It
2784Numeric types sort in numeric order.
2785.It
2786String types sort in alphabetical order.
2787.It
2788Types inappropriate for a row sort that row to the literal bottom, regardless of
2789the specified ordering.
2790.El
2791.Pp
2792If no sorting options are specified the existing behavior of
2793.Nm zfs Cm list
2794is preserved.
2795.It Fl t Ar type
2796A comma-separated list of types to display, where
2797.Ar type
2798is one of
2799.Sy filesystem ,
2800.Sy snapshot ,
2801.Sy volume ,
2802.Sy bookmark ,
2803or
2804.Sy all .
2805For example, specifying
2806.Fl t Sy snapshot
2807displays only snapshots.
2808.El
2809.It Xo
2810.Nm
2811.Cm set
2812.Ar property Ns = Ns Ar value Oo Ar property Ns = Ns Ar value Oc Ns ...
2813.Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot Ns ...
2814.Xc
2815Sets the property or list of properties to the given value(s) for each dataset.
2816Only some properties can be edited.
2817See the
2818.Sx Properties
2819section for more information on what properties can be set and acceptable
2820values.
2821Numeric values can be specified as exact values, or in a human-readable form
2822with a suffix of
2823.Sy B , K , M , G , T , P , E , Z
2824.Po for bytes, kilobytes, megabytes, gigabytes, terabytes, petabytes, exabytes,
2825or zettabytes, respectively
2826.Pc .
2827User properties can be set on snapshots.
2828For more information, see the
2829.Sx User Properties
2830section.
2831.It Xo
2832.Nm
2833.Cm get
2834.Op Fl r Ns | Ns Fl d Ar depth
2835.Op Fl Hp
2836.Oo Fl o Ar field Ns Oo , Ns Ar field Oc Ns ... Oc
2837.Oo Fl s Ar source Ns Oo , Ns Ar source Oc Ns ... Oc
2838.Oo Fl t Ar type Ns Oo , Ns Ar type Oc Ns ... Oc
2839.Cm all | Ar property Ns Oo , Ns Ar property Oc Ns ...
2840.Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot Ns | Ns Ar bookmark Ns ...
2841.Xc
2842Displays properties for the given datasets.
2843If no datasets are specified, then the command displays properties for all
2844datasets on the system.
2845For each property, the following columns are displayed:
2846.Bd -literal
2847    name      Dataset name
2848    property  Property name
2849    value     Property value
2850    source    Property source.  Can either be local, default,
2851              temporary, inherited, or none (-).
2852.Ed
2853.Pp
2854All columns are displayed by default, though this can be controlled by using the
2855.Fl o
2856option.
2857This command takes a comma-separated list of properties as described in the
2858.Sx Native Properties
2859and
2860.Sx User Properties
2861sections.
2862.Pp
2863The special value
2864.Sy all
2865can be used to display all properties that apply to the given dataset's type
2866.Pq filesystem, volume, snapshot, or bookmark .
2867.Bl -tag -width "-H"
2868.It Fl H
2869Display output in a form more easily parsed by scripts.
2870Any headers are omitted, and fields are explicitly separated by a single tab
2871instead of an arbitrary amount of space.
2872.It Fl d Ar depth
2873Recursively display any children of the dataset, limiting the recursion to
2874.Ar depth .
2875A depth of
2876.Sy 1
2877will display only the dataset and its direct children.
2878.It Fl o Ar field
2879A comma-separated list of columns to display.
2880.Sy name Ns \&, Ns Sy property Ns \&, Ns Sy value Ns \&, Ns Sy source
2881is the default value.
2882.It Fl p
2883Display numbers in parsable
2884.Pq exact
2885values.
2886.It Fl r
2887Recursively display properties for any children.
2888.It Fl s Ar source
2889A comma-separated list of sources to display.
2890Those properties coming from a source other than those in this list are ignored.
2891Each source must be one of the following:
2892.Sy local ,
2893.Sy default ,
2894.Sy inherited ,
2895.Sy temporary ,
2896and
2897.Sy none .
2898The default value is all sources.
2899.It Fl t Ar type
2900A comma-separated list of types to display, where
2901.Ar type
2902is one of
2903.Sy filesystem ,
2904.Sy snapshot ,
2905.Sy volume ,
2906.Sy bookmark ,
2907or
2908.Sy all .
2909.El
2910.It Xo
2911.Nm
2912.Cm inherit
2913.Op Fl rS
2914.Ar property Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot Ns ...
2915.Xc
2916Clears the specified property, causing it to be inherited from an ancestor,
2917restored to default if no ancestor has the property set, or with the
2918.Fl S
2919option reverted to the received value if one exists.
2920See the
2921.Sx Properties
2922section for a listing of default values, and details on which properties can be
2923inherited.
2924.Bl -tag -width "-r"
2925.It Fl r
2926Recursively inherit the given property for all children.
2927.It Fl S
2928Revert the property to the received value if one exists; otherwise operate as
2929if the
2930.Fl S
2931option was not specified.
2932.El
2933.It Xo
2934.Nm
2935.Cm remap
2936.Ar filesystem Ns | Ns Ar volume
2937.Xc
2938Remap the indirect blocks in the given filesystem or volume so that they no
2939longer reference blocks on previously removed vdevs and we can eventually
2940shrink the size of the indirect mapping objects for the previously removed
2941vdevs. Note that remapping all blocks might not be possible and that
2942references from snapshots will still exist and cannot be remapped.
2943.It Xo
2944.Nm
2945.Cm upgrade
2946.Xc
2947Displays a list of file systems that are not the most recent version.
2948.It Xo
2949.Nm
2950.Cm upgrade
2951.Fl v
2952.Xc
2953Displays a list of currently supported file system versions.
2954.It Xo
2955.Nm
2956.Cm upgrade
2957.Op Fl r
2958.Op Fl V Ar version
2959.Fl a | Ar filesystem
2960.Xc
2961Upgrades file systems to a new on-disk version.
2962Once this is done, the file systems will no longer be accessible on systems
2963running older versions of the software.
2964.Nm zfs Cm send
2965streams generated from new snapshots of these file systems cannot be accessed on
2966systems running older versions of the software.
2967.Pp
2968In general, the file system version is independent of the pool version.
2969See
2970.Xr zpool 8
2971for information on the
2972.Nm zpool Cm upgrade
2973command.
2974.Pp
2975In some cases, the file system version and the pool version are interrelated and
2976the pool version must be upgraded before the file system version can be
2977upgraded.
2978.Bl -tag -width "-V"
2979.It Fl V Ar version
2980Upgrade to the specified
2981.Ar version .
2982If the
2983.Fl V
2984flag is not specified, this command upgrades to the most recent version.
2985This
2986option can only be used to increase the version number, and only up to the most
2987recent version supported by this software.
2988.It Fl a
2989Upgrade all file systems on all imported pools.
2990.It Ar filesystem
2991Upgrade the specified file system.
2992.It Fl r
2993Upgrade the specified file system and all descendent file systems.
2994.El
2995.It Xo
2996.Nm
2997.Cm userspace
2998.Op Fl Hinp
2999.Oo Fl o Ar field Ns Oo , Ns Ar field Oc Ns ... Oc
3000.Oo Fl s Ar field Oc Ns ...
3001.Oo Fl S Ar field Oc Ns ...
3002.Oo Fl t Ar type Ns Oo , Ns Ar type Oc Ns ... Oc
3003.Ar filesystem Ns | Ns Ar snapshot
3004.Xc
3005Displays space consumed by, and quotas on, each user in the specified filesystem
3006or snapshot.
3007This corresponds to the
3008.Sy userused@ Ns Em user ,
3009.Sy userobjused@ Ns Em user ,
3010.Sy userquota@ Ns Em user,
3011and
3012.Sy userobjquota@ Ns Em user
3013properties.
3014.Bl -tag -width "-H"
3015.It Fl H
3016Do not print headers, use tab-delimited output.
3017.It Fl S Ar field
3018Sort by this field in reverse order.
3019See
3020.Fl s .
3021.It Fl i
3022Translate SID to POSIX ID.
3023The POSIX ID may be ephemeral if no mapping exists.
3024Normal POSIX interfaces
3025.Po for example,
3026.Xr stat 2 ,
3027.Nm ls Fl l
3028.Pc
3029perform this translation, so the
3030.Fl i
3031option allows the output from
3032.Nm zfs Cm userspace
3033to be compared directly with those utilities.
3034However,
3035.Fl i
3036may lead to confusion if some files were created by an SMB user before a
3037SMB-to-POSIX name mapping was established.
3038In such a case, some files will be owned by the SMB entity and some by the POSIX
3039entity.
3040However, the
3041.Fl i
3042option will report that the POSIX entity has the total usage and quota for both.
3043.It Fl n
3044Print numeric ID instead of user/group name.
3045.It Fl o Ar field Ns Oo , Ns Ar field Oc Ns ...
3046Display only the specified fields from the following set:
3047.Sy type ,
3048.Sy name ,
3049.Sy used ,
3050.Sy quota .
3051The default is to display all fields.
3052.It Fl p
3053Use exact
3054.Pq parsable
3055numeric output.
3056.It Fl s Ar field
3057Sort output by this field.
3058The
3059.Fl s
3060and
3061.Fl S
3062flags may be specified multiple times to sort first by one field, then by
3063another.
3064The default is
3065.Fl s Sy type Fl s Sy name .
3066.It Fl t Ar type Ns Oo , Ns Ar type Oc Ns ...
3067Print only the specified types from the following set:
3068.Sy all ,
3069.Sy posixuser ,
3070.Sy smbuser ,
3071.Sy posixgroup ,
3072.Sy smbgroup .
3073The default is
3074.Fl t Sy posixuser Ns \&, Ns Sy smbuser .
3075The default can be changed to include group types.
3076.El
3077.It Xo
3078.Nm
3079.Cm groupspace
3080.Op Fl Hinp
3081.Oo Fl o Ar field Ns Oo , Ns Ar field Oc Ns ... Oc
3082.Oo Fl s Ar field Oc Ns ...
3083.Oo Fl S Ar field Oc Ns ...
3084.Oo Fl t Ar type Ns Oo , Ns Ar type Oc Ns ... Oc
3085.Ar filesystem Ns | Ns Ar snapshot
3086.Xc
3087Displays space consumed by, and quotas on, each group in the specified
3088filesystem or snapshot.
3089This subcommand is identical to
3090.Nm zfs Cm userspace ,
3091except that the default types to display are
3092.Fl t Sy posixgroup Ns \&, Ns Sy smbgroup .
3093.It Xo
3094.Nm
3095.Cm projectspace
3096.Op Fl Hp
3097.Oo Fl o Ar field Ns Oo , Ns Ar field Oc Ns ... Oc
3098.Oo Fl s Ar field Oc Ns ...
3099.Oo Fl S Ar field Oc Ns ...
3100.Ar filesystem Ns | Ns Ar snapshot
3101.Xc
3102Displays space consumed by, and quotas on, each project in the specified
3103filesystem or snapshot.
3104This subcommand is identical to
3105.Nm zfs Cm userspace ,
3106except that the project identifier is numeral, not name.
3107So need neither the option
3108.Sy -i
3109for SID to POSIX ID nor
3110.Sy -n
3111for numeric ID, nor
3112.Sy -t
3113for types.
3114.It Xo
3115.Nm
3116.Cm project
3117.Oo Fl d Ns | Ns Fl r Ns Oc
3118.Ar file Ns | Ns Ar directory Ns ...
3119.Xc
3120List project identifier (ID) and inherit flag of files or directories.
3121.Bl -tag -width "-d"
3122.It Fl d
3123Show the directory project ID and inherit flag, not its children.
3124It will overwrite the former specified
3125.Fl r
3126option.
3127.It Fl r
3128Show on subdirectories recursively.
3129It will overwrite the former specified
3130.Fl d
3131option.
3132.El
3133.It Xo
3134.Nm
3135.Cm project
3136.Fl C
3137.Oo Fl kr Ns Oc
3138.Ar file Ns | Ns Ar directory Ns ...
3139.Xc
3140Clear project inherit flag and/or ID on the files or directories.
3141.Bl -tag -width "-k"
3142.It Fl k
3143Keep the project ID unchanged.
3144If not specified, the project ID will be reset as zero.
3145.It Fl r
3146Clear on subdirectories recursively.
3147.El
3148.It Xo
3149.Nm
3150.Cm project
3151.Fl c
3152.Oo Fl 0 Ns Oc
3153.Oo Fl d Ns | Ns Fl r Ns Oc
3154.Op Fl p Ar id
3155.Ar file Ns | Ns Ar directory Ns ...
3156.Xc
3157Check project ID and inherit flag on the files or directories, report the
3158entries without project inherit flag or with different project IDs from the
3159specified (via
3160.Fl p
3161option) value or the target directory's project ID.
3162.Bl -tag -width "-0"
3163.It Fl 0
3164Print file name with a trailing NUL instead of newline (by default), like
3165"find -print0".
3166.It Fl d
3167Check the directory project ID and inherit flag, not its children.
3168It will overwrite the former specified
3169.Fl r
3170option.
3171.It Fl p
3172Specify the referenced ID for comparing with the target files or directories'
3173project IDs.
3174If not specified, the target (top) directory's project ID will be used as the
3175referenced one.
3176.It Fl r
3177Check on subdirectories recursively.
3178It will overwrite the former specified
3179.Fl d
3180option.
3181.El
3182.It Xo
3183.Nm
3184.Cm project
3185.Op Fl p Ar id
3186.Oo Fl rs Ns Oc
3187.Ar file Ns | Ns Ar directory Ns ...
3188.Xc
3189Set project ID and/or inherit flag on the files or directories.
3190.Bl -tag -width "-p"
3191.It Fl p
3192Set the files' or directories' project ID with the given value.
3193.It Fl r
3194Set on subdirectories recursively.
3195.It Fl s
3196Set project inherit flag on the given files or directories.
3197It is usually used for setup tree quota on the directory target with
3198.Fl r
3199option specified together.
3200When setup tree quota, by default the directory's project ID will be set to
3201all its descendants unless you specify the project ID via
3202.Fl p
3203option explicitly.
3204.El
3205.It Xo
3206.Nm
3207.Cm mount
3208.Xc
3209Displays all ZFS file systems currently mounted.
3210.It Xo
3211.Nm
3212.Cm mount
3213.Op Fl Olv
3214.Op Fl o Ar options
3215.Fl a | Ar filesystem
3216.Xc
3217Mounts ZFS file systems.
3218.Bl -tag -width "-O"
3219.It Fl O
3220Perform an overlay mount.
3221See
3222.Xr mount 8
3223for more information.
3224.It Fl a
3225Mount all available ZFS file systems.
3226Invoked automatically as part of the boot process.
3227.It Fl l
3228Load keys for encrypted filesystems as they are being mounted.
3229This is equivalent to executing
3230.Nm Cm load-key
3231on each encryption root before mounting it.
3232Note that if a filesystem has a
3233.Sy keylocation
3234of
3235.Sy prompt
3236this will cause the terminal to interactively block after asking for the key.
3237.It Ar filesystem
3238Mount the specified filesystem.
3239.It Fl o Ar options
3240An optional, comma-separated list of mount options to use temporarily for the
3241duration of the mount.
3242See the
3243.Sx Temporary Mount Point Properties
3244section for details.
3245.It Fl v
3246Report mount progress.
3247.El
3248.It Xo
3249.Nm
3250.Cm unmount
3251.Op Fl f
3252.Fl a | Ar filesystem Ns | Ns Ar mountpoint
3253.Xc
3254Unmounts currently mounted ZFS file systems.
3255.Bl -tag -width "-a"
3256.It Fl a
3257Unmount all available ZFS file systems.
3258Invoked automatically as part of the shutdown process.
3259.It Ar filesystem Ns | Ns Ar mountpoint
3260Unmount the specified filesystem.
3261The command can also be given a path to a ZFS file system mount point on the
3262system.
3263.It Fl f
3264Forcefully unmount the file system, even if it is currently in use.
3265.El
3266.It Xo
3267.Nm
3268.Cm share
3269.Fl a | Ar filesystem
3270.Xc
3271Shares available ZFS file systems.
3272.Bl -tag -width "-a"
3273.It Fl a
3274Share all available ZFS file systems.
3275Invoked automatically as part of the boot process.
3276.It Ar filesystem
3277Share the specified filesystem according to the
3278.Sy sharenfs
3279and
3280.Sy sharesmb
3281properties.
3282File systems are shared when the
3283.Sy sharenfs
3284or
3285.Sy sharesmb
3286property is set.
3287.El
3288.It Xo
3289.Nm
3290.Cm unshare
3291.Fl a | Ar filesystem Ns | Ns Ar mountpoint
3292.Xc
3293Unshares currently shared ZFS file systems.
3294.Bl -tag -width "-a"
3295.It Fl a
3296Unshare all available ZFS file systems.
3297Invoked automatically as part of the shutdown process.
3298.It Ar filesystem Ns | Ns Ar mountpoint
3299Unshare the specified filesystem.
3300The command can also be given a path to a ZFS file system shared on the system.
3301.El
3302.It Xo
3303.Nm
3304.Cm bookmark
3305.Ar snapshot bookmark
3306.Xc
3307Creates a bookmark of the given snapshot.
3308Bookmarks mark the point in time when the snapshot was created, and can be used
3309as the incremental source for a
3310.Nm zfs Cm send
3311command.
3312.Pp
3313This feature must be enabled to be used.
3314See
3315.Xr zpool-features 7
3316for details on ZFS feature flags and the
3317.Sy bookmarks
3318feature.
3319.It Xo
3320.Nm
3321.Cm send
3322.Op Fl DLPRbcehnpvw
3323.Op Oo Fl I Ns | Ns Fl i Oc Ar snapshot
3324.Ar snapshot
3325.Xc
3326Creates a stream representation of the second
3327.Ar snapshot ,
3328which is written to standard output.
3329The output can be redirected to a file or to a different system
3330.Po for example, using
3331.Xr ssh 1
3332.Pc .
3333By default, a full stream is generated.
3334.Bl -tag -width "-D"
3335.It Fl D , -dedup
3336Generate a deduplicated stream.
3337Blocks which would have been sent multiple times in the send stream will only be
3338sent once.
3339The receiving system must also support this feature to receive a deduplicated
3340stream.
3341This flag can be used regardless of the dataset's
3342.Sy dedup
3343property, but performance will be much better if the filesystem uses a
3344dedup-capable checksum
3345.Po for example,
3346.Sy sha256
3347.Pc .
3348.It Fl I Ar snapshot
3349Generate a stream package that sends all intermediary snapshots from the first
3350snapshot to the second snapshot.
3351For example,
3352.Fl I Em @a Em fs@d
3353is similar to
3354.Fl i Em @a Em fs@b Ns \&; Fl i Em @b Em fs@c Ns \&; Fl i Em @c Em fs@d .
3355The incremental source may be specified as with the
3356.Fl i
3357option.
3358.It Fl L , -large-block
3359Generate a stream which may contain blocks larger than 128KB.
3360This flag has no effect if the
3361.Sy large_blocks
3362pool feature is disabled, or if the
3363.Sy recordsize
3364property of this filesystem has never been set above 128KB.
3365The receiving system must have the
3366.Sy large_blocks
3367pool feature enabled as well.
3368See
3369.Xr zpool-features 7
3370for details on ZFS feature flags and the
3371.Sy large_blocks
3372feature.
3373.It Fl P , -parsable
3374Print machine-parsable verbose information about the stream package generated.
3375.It Fl R , -replicate
3376Generate a replication stream package, which will replicate the specified
3377file system, and all descendent file systems, up to the named snapshot.
3378When received, all properties, snapshots, descendent file systems, and clones
3379are preserved.
3380.Pp
3381If the
3382.Fl i
3383or
3384.Fl I
3385flags are used in conjunction with the
3386.Fl R
3387flag, an incremental replication stream is generated.
3388The current values of properties, and current snapshot and file system names are
3389set when the stream is received.
3390If the
3391.Fl F
3392flag is specified when this stream is received, snapshots and file systems that
3393do not exist on the sending side are destroyed.
3394If the
3395.Fl R
3396flag is used to send encrypted datasets, then
3397.Fl w
3398must also be specified.
3399.It Fl e , -embed
3400Generate a more compact stream by using
3401.Sy WRITE_EMBEDDED
3402records for blocks which are stored more compactly on disk by the
3403.Sy embedded_data
3404pool feature.
3405This flag has no effect if the
3406.Sy embedded_data
3407feature is disabled.
3408The receiving system must have the
3409.Sy embedded_data
3410feature enabled.
3411If the
3412.Sy lz4_compress
3413feature is active on the sending system, then the receiving system must have
3414that feature enabled as well.
3415Datasets that are sent with this flag may not be received as an encrypted
3416dataset, since encrypted datasets cannot use the
3417.Sy embedded_data
3418feature.
3419See
3420.Xr zpool-features 7
3421for details on ZFS feature flags and the
3422.Sy embedded_data
3423feature.
3424.It Fl b, -backup
3425Sends only received property values whether or not they are overridden by local
3426settings, but only if the dataset has ever been received.
3427Use this option when you want
3428.Nm zfs Cm receive
3429to restore received properties backed up on the sent dataset and to avoid
3430sending local settings that may have nothing to do with the source dataset,
3431but only with how the data is backed up.
3432.It Fl c , -compressed
3433Generate a more compact stream by using compressed WRITE records for blocks
3434which are compressed on disk and in memory
3435.Po see the
3436.Sy compression
3437property for details
3438.Pc .
3439If the
3440.Sy lz4_compress
3441feature is active on the sending system, then the receiving system must have
3442that feature enabled as well.
3443If the
3444.Sy large_blocks
3445feature is enabled on the sending system but the
3446.Fl L
3447option is not supplied in conjunction with
3448.Fl c ,
3449then the data will be decompressed before sending so it can be split into
3450smaller block sizes.
3451.It Fl h, -holds
3452Generate a stream package that includes any snapshot holds (created with the
3453.Sy zfs hold
3454command), and indicating to
3455.Sy zfs receive
3456that the holds be applied to the dataset on the receiving system.
3457.It Fl i Ar snapshot
3458Generate an incremental stream from the first
3459.Ar snapshot
3460.Pq the incremental source
3461to the second
3462.Ar snapshot
3463.Pq the incremental target .
3464The incremental source can be specified as the last component of the snapshot
3465name
3466.Po the
3467.Sy @
3468character and following
3469.Pc
3470and it is assumed to be from the same file system as the incremental target.
3471.Pp
3472If the destination is a clone, the source may be the origin snapshot, which must
3473be fully specified
3474.Po for example,
3475.Em pool/fs@origin ,
3476not just
3477.Em @origin
3478.Pc .
3479.It Fl n , -dryrun
3480Do a dry-run
3481.Pq Qq No-op
3482send.
3483Do not generate any actual send data.
3484This is useful in conjunction with the
3485.Fl v
3486or
3487.Fl P
3488flags to determine what data will be sent.
3489In this case, the verbose output will be written to standard output
3490.Po contrast with a non-dry-run, where the stream is written to standard output
3491and the verbose output goes to standard error
3492.Pc .
3493.It Fl p , -props
3494Include the dataset's properties in the stream.
3495This flag is implicit when
3496.Fl R
3497is specified.
3498The receiving system must also support this feature.
3499Sends of encrypted datasets must use
3500.Fl w
3501when using this flag.
3502.It Fl w , -raw
3503For encrypted datasets, send data exactly as it exists on disk.
3504This allows backups to be taken even if encryption keys are not currently
3505loaded.
3506The backup may then be received on an untrusted machine since that machine will
3507not have the encryption keys to read the protected data or alter it without
3508being detected.
3509Upon being received, the dataset will have the same encryption keys as it did
3510on the send side, although the
3511.Sy keylocation
3512property will be defaulted to
3513.Sy prompt
3514if not otherwise provided.
3515For unencrypted datasets, this flag will be equivalent to
3516.Fl Lec .
3517Note that if you do not use this flag for sending encrypted datasets,
3518data will be sent unencrypted and may be re-encrypted with a different
3519encryption key on the receiving system, which will disable the ability
3520to do a raw send to that system for incrementals.
3521.It Fl v , -verbose
3522Print verbose information about the stream package generated.
3523This information includes a per-second report of how much data has been sent.
3524.Pp
3525The format of the stream is committed.
3526You will be able to receive your streams on future versions of ZFS .
3527.El
3528.It Xo
3529.Nm
3530.Cm send
3531.Op Fl Lcew
3532.Op Fl i Ar snapshot Ns | Ns Ar bookmark
3533.Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot
3534.Xc
3535Generate a send stream, which may be of a filesystem, and may be incremental
3536from a bookmark.
3537If the destination is a filesystem or volume, the pool must be read-only, or the
3538filesystem must not be mounted.
3539When the stream generated from a filesystem or volume is received, the default
3540snapshot name will be
3541.Qq --head-- .
3542.Bl -tag -width "-L"
3543.It Fl L , -large-block
3544Generate a stream which may contain blocks larger than 128KB.
3545This flag has no effect if the
3546.Sy large_blocks
3547pool feature is disabled, or if the
3548.Sy recordsize
3549property of this filesystem has never been set above 128KB.
3550The receiving system must have the
3551.Sy large_blocks
3552pool feature enabled as well.
3553See
3554.Xr zpool-features 7
3555for details on ZFS feature flags and the
3556.Sy large_blocks
3557feature.
3558.It Fl c , -compressed
3559Generate a more compact stream by using compressed WRITE records for blocks
3560which are compressed on disk and in memory
3561.Po see the
3562.Sy compression
3563property for details
3564.Pc .
3565If the
3566.Sy lz4_compress
3567feature is active on the sending system, then the receiving system must have
3568that feature enabled as well.
3569If the
3570.Sy large_blocks
3571feature is enabled on the sending system but the
3572.Fl L
3573option is not supplied in conjunction with
3574.Fl c ,
3575then the data will be decompressed before sending so it can be split into
3576smaller block sizes.
3577.It Fl e , -embed
3578Generate a more compact stream by using
3579.Sy WRITE_EMBEDDED
3580records for blocks which are stored more compactly on disk by the
3581.Sy embedded_data
3582pool feature.
3583This flag has no effect if the
3584.Sy embedded_data
3585feature is disabled.
3586The receiving system must have the
3587.Sy embedded_data
3588feature enabled.
3589If the
3590.Sy lz4_compress
3591feature is active on the sending system, then the receiving system must have
3592that feature enabled as well.
3593Datasets that are sent with this flag may not be received as an encrypted
3594dataset, since encrypted datasets cannot use the
3595.Sy embedded_data
3596feature.
3597See
3598.Xr zpool-features 7
3599for details on ZFS feature flags and the
3600.Sy embedded_data
3601feature.
3602.It Fl i Ar snapshot Ns | Ns Ar bookmark
3603Generate an incremental send stream.
3604The incremental source must be an earlier snapshot in the destination's history.
3605It will commonly be an earlier snapshot in the destination's file system, in
3606which case it can be specified as the last component of the name
3607.Po the
3608.Sy #
3609or
3610.Sy @
3611character and following
3612.Pc .
3613.Pp
3614If the incremental target is a clone, the incremental source can be the origin
3615snapshot, or an earlier snapshot in the origin's filesystem, or the origin's
3616origin, etc.
3617.It Fl w , -raw
3618For encrypted datasets, send data exactly as it exists on disk.
3619This allows backups to be taken even if encryption keys are not currently
3620loaded.
3621The backup may then be received on an untrusted machine since that machine will
3622not have the encryption keys to read the protected data or alter it without
3623being detected.
3624Upon being received, the dataset will have the same encryption keys as it did
3625on the send side, although the
3626.Sy keylocation
3627property will be defaulted to
3628.Sy prompt
3629if not otherwise provided.
3630For unencrypted datasets, this flag will be equivalent to
3631.Fl Lec .
3632Note that if you do not use this flag for sending encrypted datasets,
3633data will be sent unencrypted and may be re-encrypted with a different
3634encryption key on the receiving system, which will disable the ability
3635to do a raw send to that system for incrementals.
3636.El
3637.It Xo
3638.Nm
3639.Cm send
3640.Op Fl Penv
3641.Fl t
3642.Ar receive_resume_token
3643.Xc
3644Creates a send stream which resumes an interrupted receive.
3645The
3646.Ar receive_resume_token
3647is the value of this property on the filesystem or volume that was being
3648received into.
3649See the documentation for
3650.Sy zfs receive -s
3651for more details.
3652.It Xo
3653.Nm
3654.Cm receive
3655.Op Fl Fhnsuv
3656.Op Fl o Sy origin Ns = Ns Ar snapshot
3657.Op Fl o Ar property Ns = Ns Ar value
3658.Op Fl x Ar property
3659.Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot
3660.Xc
3661.It Xo
3662.Nm
3663.Cm receive
3664.Op Fl Fhnsuv
3665.Op Fl d Ns | Ns Fl e
3666.Op Fl o Sy origin Ns = Ns Ar snapshot
3667.Op Fl o Ar property Ns = Ns Ar value
3668.Op Fl x Ar property
3669.Ar filesystem
3670.Xc
3671Creates a snapshot whose contents are as specified in the stream provided on
3672standard input.
3673If a full stream is received, then a new file system is created as well.
3674Streams are created using the
3675.Nm zfs Cm send
3676subcommand, which by default creates a full stream.
3677.Nm zfs Cm recv
3678can be used as an alias for
3679.Nm zfs Cm receive .
3680.Pp
3681If an incremental stream is received, then the destination file system must
3682already exist, and its most recent snapshot must match the incremental stream's
3683source.
3684For
3685.Sy zvols ,
3686the destination device link is destroyed and recreated, which means the
3687.Sy zvol
3688cannot be accessed during the
3689.Cm receive
3690operation.
3691.Pp
3692When a snapshot replication package stream that is generated by using the
3693.Nm zfs Cm send Fl R
3694command is received, any snapshots that do not exist on the sending location are
3695destroyed by using the
3696.Nm zfs Cm destroy Fl d
3697command.
3698.Pp
3699If
3700.Fl o Em property Ns = Ns Ar value
3701or
3702.Fl x Em property
3703is specified, it applies to the effective value of the property throughout
3704the entire subtree of replicated datasets.
3705Effective property values will be
3706set (
3707.Fl o
3708) or inherited (
3709.Fl x
3710) on the topmost in the replicated subtree.
3711In descendant datasets, if the property is set by the send stream, it will be
3712overridden by forcing the property to be inherited from the top‐most file
3713system.
3714Received properties are retained in spite of being overridden and may be
3715restored with
3716.Nm zfs Cm inherit Fl S .
3717Specifying
3718.Fl o Sy origin Ns = Ns Em snapshot
3719is a special case because, even if
3720.Sy origin
3721is a read-only property and cannot be set, it's allowed to receive the send
3722stream as a clone of the given snapshot.
3723.Pp
3724Raw encrypted send streams (created with
3725.Nm zfs Cm send Fl w
3726) may only be received as is, and cannot be re-encrypted, decrypted, or
3727recompressed by the receive process.
3728Unencrypted streams can be received as encrypted datasets, either through
3729inheritance or by specifying encryption parameters with the
3730.Fl o
3731options.
3732Note that the
3733.Sy keylocation
3734property cannot be overridden to
3735.Sy prompt
3736during a receive.
3737This is because the receive process itself is already using
3738stdin for the send stream.
3739Instead, the property can be overridden after the receive completes.
3740.Pp
3741The added security provided by raw sends adds some restrictions to the send
3742and receive process.
3743ZFS will not allow a mix of raw receives and non-raw receives.
3744Specifically, any raw incremental receives that are attempted after
3745a non-raw receive will fail.
3746Non-raw receives do not have this restriction and, therefore, are always
3747possible.
3748Because of this, it is best practice to always use either raw sends for
3749their security benefits or non-raw sends for their flexibility when working
3750with encrypted datasets, but not a combination.
3751.Pp
3752The reason for this restriction stems from the inherent restrictions of the
3753AEAD ciphers that ZFS uses to encrypt data.
3754When using ZFS native encryption, each block of data is encrypted against
3755a randomly generated number known as the "initialization vector" (IV),
3756which is stored in the filesystem metadata.
3757This number is required by the encryption algorithms whenever the data is to
3758be decrypted.
3759Together, all of the IVs provided for all of the blocks in a given snapshot
3760are collectively called an "IV set".
3761When ZFS performs a raw send, the IV set is transferred from the source to
3762the destination in the send stream.
3763When ZFS performs a non-raw send, the data is decrypted by the source
3764system and re-encrypted by the destination system, creating a snapshot with
3765effectively the same data, but a different IV set.
3766In order for decryption to work after a raw send, ZFS must ensure that the
3767IV set used on both the source and destination side match.
3768When an incremental raw receive is performed on top of an existing snapshot,
3769ZFS will check to confirm that the "from" snapshot on both the source and
3770destination were using the same IV set, ensuring the new IV set is consistent.
3771.Pp
3772The name of the snapshot
3773.Pq and file system, if a full stream is received
3774that this subcommand creates depends on the argument type and the use of the
3775.Fl d
3776or
3777.Fl e
3778options.
3779.Pp
3780If the argument is a snapshot name, the specified
3781.Ar snapshot
3782is created.
3783If the argument is a file system or volume name, a snapshot with the same name
3784as the sent snapshot is created within the specified
3785.Ar filesystem
3786or
3787.Ar volume .
3788If neither of the
3789.Fl d
3790or
3791.Fl e
3792options are specified, the provided target snapshot name is used exactly as
3793provided.
3794.Pp
3795The
3796.Fl d
3797and
3798.Fl e
3799options cause the file system name of the target snapshot to be determined by
3800appending a portion of the sent snapshot's name to the specified target
3801.Ar filesystem .
3802If the
3803.Fl d
3804option is specified, all but the first element of the sent snapshot's file
3805system path
3806.Pq usually the pool name
3807is used and any required intermediate file systems within the specified one are
3808created.
3809If the
3810.Fl e
3811option is specified, then only the last element of the sent snapshot's file
3812system name
3813.Pq i.e. the name of the source file system itself
3814is used as the target file system name.
3815.Bl -tag -width "-F"
3816.It Fl F
3817Force a rollback of the file system to the most recent snapshot before
3818performing the receive operation.
3819If receiving an incremental replication stream
3820.Po for example, one generated by
3821.Nm zfs Cm send Fl R Op Fl i Ns | Ns Fl I
3822.Pc ,
3823destroy snapshots and file systems that do not exist on the sending side.
3824.It Fl d
3825Discard the first element of the sent snapshot's file system name, using the
3826remaining elements to determine the name of the target file system for the new
3827snapshot as described in the paragraph above.
3828.It Fl e
3829Discard all but the last element of the sent snapshot's file system name, using
3830that element to determine the name of the target file system for the new
3831snapshot as described in the paragraph above.
3832.It Fl h
3833Skip the receive of holds.
3834There is no effect if holds are not sent.
3835.It Fl n
3836Do not actually receive the stream.
3837This can be useful in conjunction with the
3838.Fl v
3839option to verify the name the receive operation would use.
3840.It Fl o Sy origin Ns = Ns Ar snapshot
3841Forces the stream to be received as a clone of the given snapshot.
3842If the stream is a full send stream, this will create the filesystem
3843described by the stream as a clone of the specified snapshot.
3844Which snapshot was specified will not affect the success or failure of the
3845receive, as long as the snapshot does exist.
3846If the stream is an incremental send stream, all the normal verification will be
3847performed.
3848.It Fl o Em property Ns = Ns Ar value
3849Sets the specified property as if the command
3850.Nm zfs Cm set Em property Ns = Ns Ar value
3851was invoked immediately before the receive.
3852When receiving a stream from
3853.Nm zfs Cm send Fl R ,
3854causes the property to be inherited by all descendant datasets, as though
3855.Nm zfs Cm inherit Em property
3856was run on any descendant datasets that have this property set on the
3857sending system.
3858.Pp
3859Any editable property can be set at receive time.
3860Set-once properties bound to the received data, such as
3861.Sy normalization
3862and
3863.Sy casesensitivity ,
3864cannot be set at receive time even when the datasets are newly created by
3865.Nm zfs Cm receive .
3866Additionally both settable properties
3867.Sy version
3868and
3869.Sy volsize
3870cannot be set at receive time.
3871.Pp
3872The
3873.Fl o
3874option may be specified multiple times, for different properties.
3875An error results if the same property is specified in multiple
3876.Fl o
3877or
3878.Fl x
3879options.
3880.Pp
3881The
3882.Fl o
3883option may also be used to override encryption properties upon initial
3884receive.
3885This allows unencrypted streams to be received as encrypted datasets.
3886To cause the received dataset (or root dataset of a recursive stream) to be
3887received as an encryption root, specify encryption properties in the same
3888manner as is required for
3889.Nm
3890.Cm create .
3891For instance:
3892.Bd -literal
3893# zfs send tank/test@snap1 | zfs recv -o encryption=on -o keyformat=passphrase -o keylocation=file:///path/to/keyfile
3894.Ed
3895.Pp
3896Note that
3897.Op Fl o Ar keylocation Ns = Ns Ar prompt
3898may not be specified here, since stdin is already being utilized for the send
3899stream.
3900Once the receive has completed, you can use
3901.Nm
3902.Cm set
3903to change this setting after the fact.
3904Similarly, you can receive a dataset as an encrypted child by specifying
3905.Op Fl x Ar encryption
3906to force the property to be inherited.
3907Overriding encryption properties (except for
3908.Sy keylocation )
3909is not possible with raw send streams.
3910.It Fl s
3911If the receive is interrupted, save the partially received state, rather
3912than deleting it.
3913Interruption may be due to premature termination of the stream
3914.Po e.g. due to network failure or failure of the remote system
3915if the stream is being read over a network connection
3916.Pc ,
3917a checksum error in the stream, termination of the
3918.Nm zfs Cm receive
3919process, or unclean shutdown of the system.
3920.Pp
3921The receive can be resumed with a stream generated by
3922.Nm zfs Cm send Fl t Ar token ,
3923where the
3924.Ar token
3925is the value of the
3926.Sy receive_resume_token
3927property of the filesystem or volume which is received into.
3928.Pp
3929To use this flag, the storage pool must have the
3930.Sy extensible_dataset
3931feature enabled.
3932See
3933.Xr zpool-features 7
3934for details on ZFS feature flags.
3935.It Fl u
3936File system that is associated with the received stream is not mounted.
3937.It Fl v
3938Print verbose information about the stream and the time required to perform the
3939receive operation.
3940.It Fl x Em property
3941Ensures that the effective value of the specified property after the
3942receive is unaffected by the value of that property in the send stream (if any),
3943as if the property had been excluded from the send stream.
3944.Pp
3945If the specified property is not present in the send stream, this option does
3946nothing.
3947.Pp
3948If a received property needs to be overridden, the effective value will be
3949set or inherited, depending on whether the property is inheritable or not.
3950.Pp
3951In the case of an incremental update,
3952.Fl x
3953leaves any existing local setting or explicit inheritance unchanged.
3954.Pp
3955All
3956.Fl o
3957restrictions (e.g. set-once) apply equally to
3958.Fl x .
3959.El
3960.It Xo
3961.Nm
3962.Cm receive
3963.Fl A
3964.Ar filesystem Ns | Ns Ar volume
3965.Xc
3966Abort an interrupted
3967.Nm zfs Cm receive Fl s ,
3968deleting its saved partially received state.
3969.It Xo
3970.Nm
3971.Cm allow
3972.Ar filesystem Ns | Ns Ar volume
3973.Xc
3974Displays permissions that have been delegated on the specified filesystem or
3975volume.
3976See the other forms of
3977.Nm zfs Cm allow
3978for more information.
3979.It Xo
3980.Nm
3981.Cm allow
3982.Op Fl dglu
3983.Ar user Ns | Ns Ar group Ns Oo , Ns Ar user Ns | Ns Ar group Oc Ns ...
3984.Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
3985.Ar setname Oc Ns ...
3986.Ar filesystem Ns | Ns Ar volume
3987.Xc
3988.It Xo
3989.Nm
3990.Cm allow
3991.Op Fl dl
3992.Fl e Ns | Ns Sy everyone
3993.Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
3994.Ar setname Oc Ns ...
3995.Ar filesystem Ns | Ns Ar volume
3996.Xc
3997Delegates ZFS administration permission for the file systems to non-privileged
3998users.
3999.Bl -tag -width "-d"
4000.It Fl d
4001Allow only for the descendent file systems.
4002.It Fl e Ns | Ns Sy everyone
4003Specifies that the permissions be delegated to everyone.
4004.It Fl g Ar group Ns Oo , Ns Ar group Oc Ns ...
4005Explicitly specify that permissions are delegated to the group.
4006.It Fl l
4007Allow
4008.Qq locally
4009only for the specified file system.
4010.It Fl u Ar user Ns Oo , Ns Ar user Oc Ns ...
4011Explicitly specify that permissions are delegated to the user.
4012.It Ar user Ns | Ns Ar group Ns Oo , Ns Ar user Ns | Ns Ar group Oc Ns ...
4013Specifies to whom the permissions are delegated.
4014Multiple entities can be specified as a comma-separated list.
4015If neither of the
4016.Fl gu
4017options are specified, then the argument is interpreted preferentially as the
4018keyword
4019.Sy everyone ,
4020then as a user name, and lastly as a group name.
4021To specify a user or group named
4022.Qq everyone ,
4023use the
4024.Fl g
4025or
4026.Fl u
4027options.
4028To specify a group with the same name as a user, use the
4029.Fl g
4030options.
4031.It Xo
4032.Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
4033.Ar setname Oc Ns ...
4034.Xc
4035The permissions to delegate.
4036Multiple permissions may be specified as a comma-separated list.
4037Permission names are the same as ZFS subcommand and property names.
4038See the property list below.
4039Property set names, which begin with
4040.Sy @ ,
4041may be specified.
4042See the
4043.Fl s
4044form below for details.
4045.El
4046.Pp
4047If neither of the
4048.Fl dl
4049options are specified, or both are, then the permissions are allowed for the
4050file system or volume, and all of its descendents.
4051.Pp
4052Permissions are generally the ability to use a ZFS subcommand or change a ZFS
4053property.
4054The following permissions are available:
4055.Bd -literal
4056NAME             TYPE           NOTES
4057allow            subcommand     Must also have the permission that is
4058                                being allowed
4059clone            subcommand     Must also have the 'create' ability and
4060                                'mount' ability in the origin file system
4061create           subcommand     Must also have the 'mount' ability
4062destroy          subcommand     Must also have the 'mount' ability
4063diff             subcommand     Allows lookup of paths within a dataset
4064                                given an object number, and the ability
4065                                to create snapshots necessary to
4066                                'zfs diff'.
4067load-key         subcommand     Allows loading and unloading of encryption key
4068                                (see 'zfs load-key' and 'zfs unload-key').
4069change-key       subcommand     Allows changing an encryption key via
4070                                'zfs change-key'.
4071mount            subcommand     Allows mount/umount of ZFS datasets
4072promote          subcommand     Must also have the 'mount' and 'promote'
4073                                ability in the origin file system
4074receive          subcommand     Must also have the 'mount' and 'create'
4075                                ability
4076rename           subcommand     Must also have the 'mount' and 'create'
4077                                ability in the new parent
4078rollback         subcommand     Must also have the 'mount' ability
4079send             subcommand
4080share            subcommand     Allows sharing file systems over NFS
4081                                or SMB protocols
4082snapshot         subcommand     Must also have the 'mount' ability
4083
4084groupquota       other          Allows accessing any groupquota@...
4085                                property
4086groupused        other          Allows reading any groupused@... property
4087userprop         other          Allows changing any user property
4088userquota        other          Allows accessing any userquota@...
4089                                property
4090userused         other          Allows reading any userused@... property
4091projectobjquota  other          Allows accessing any projectobjquota@...
4092                                property
4093projectquota     other          Allows accessing any projectquota@... property
4094projectobjused   other          Allows reading any projectobjused@... property
4095projectused      other          Allows reading any projectused@... property
4096
4097aclinherit       property
4098aclmode          property
4099atime            property
4100canmount         property
4101casesensitivity  property
4102checksum         property
4103compression      property
4104copies           property
4105devices          property
4106exec             property
4107filesystem_limit property
4108mountpoint       property
4109nbmand           property
4110normalization    property
4111primarycache     property
4112quota            property
4113readonly         property
4114recordsize       property
4115refquota         property
4116refreservation   property
4117reservation      property
4118secondarycache   property
4119setuid           property
4120sharenfs         property
4121sharesmb         property
4122snapdir          property
4123snapshot_limit   property
4124utf8only         property
4125version          property
4126volblocksize     property
4127volsize          property
4128vscan            property
4129xattr            property
4130zoned            property
4131.Ed
4132.It Xo
4133.Nm
4134.Cm allow
4135.Fl c
4136.Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
4137.Ar setname Oc Ns ...
4138.Ar filesystem Ns | Ns Ar volume
4139.Xc
4140Sets
4141.Qq create time
4142permissions.
4143These permissions are granted
4144.Pq locally
4145to the creator of any newly-created descendent file system.
4146.It Xo
4147.Nm
4148.Cm allow
4149.Fl s No @ Ns Ar setname
4150.Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
4151.Ar setname Oc Ns ...
4152.Ar filesystem Ns | Ns Ar volume
4153.Xc
4154Defines or adds permissions to a permission set.
4155The set can be used by other
4156.Nm zfs Cm allow
4157commands for the specified file system and its descendents.
4158Sets are evaluated dynamically, so changes to a set are immediately reflected.
4159Permission sets follow the same naming restrictions as ZFS file systems, but the
4160name must begin with
4161.Sy @ ,
4162and can be no more than 64 characters long.
4163.It Xo
4164.Nm
4165.Cm unallow
4166.Op Fl dglru
4167.Ar user Ns | Ns Ar group Ns Oo , Ns Ar user Ns | Ns Ar group Oc Ns ...
4168.Oo Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
4169.Ar setname Oc Ns ... Oc
4170.Ar filesystem Ns | Ns Ar volume
4171.Xc
4172.It Xo
4173.Nm
4174.Cm unallow
4175.Op Fl dlr
4176.Fl e Ns | Ns Sy everyone
4177.Oo Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
4178.Ar setname Oc Ns ... Oc
4179.Ar filesystem Ns | Ns Ar volume
4180.Xc
4181.It Xo
4182.Nm
4183.Cm unallow
4184.Op Fl r
4185.Fl c
4186.Oo Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
4187.Ar setname Oc Ns ... Oc
4188.Ar filesystem Ns | Ns Ar volume
4189.Xc
4190Removes permissions that were granted with the
4191.Nm zfs Cm allow
4192command.
4193No permissions are explicitly denied, so other permissions granted are still in
4194effect.
4195For example, if the permission is granted by an ancestor.
4196If no permissions are specified, then all permissions for the specified
4197.Ar user ,
4198.Ar group ,
4199or
4200.Sy everyone
4201are removed.
4202Specifying
4203.Sy everyone
4204.Po or using the
4205.Fl e
4206option
4207.Pc
4208only removes the permissions that were granted to everyone, not all permissions
4209for every user and group.
4210See the
4211.Nm zfs Cm allow
4212command for a description of the
4213.Fl ldugec
4214options.
4215.Bl -tag -width "-r"
4216.It Fl r
4217Recursively remove the permissions from this file system and all descendents.
4218.El
4219.It Xo
4220.Nm
4221.Cm unallow
4222.Op Fl r
4223.Fl s No @ Ns Ar setname
4224.Oo Ar perm Ns | Ns @ Ns Ar setname Ns Oo , Ns Ar perm Ns | Ns @ Ns
4225.Ar setname Oc Ns ... Oc
4226.Ar filesystem Ns | Ns Ar volume
4227.Xc
4228Removes permissions from a permission set.
4229If no permissions are specified, then all permissions are removed, thus removing
4230the set entirely.
4231.It Xo
4232.Nm
4233.Cm hold
4234.Op Fl r
4235.Ar tag Ar snapshot Ns ...
4236.Xc
4237Adds a single reference, named with the
4238.Ar tag
4239argument, to the specified snapshot or snapshots.
4240Each snapshot has its own tag namespace, and tags must be unique within that
4241space.
4242.Pp
4243If a hold exists on a snapshot, attempts to destroy that snapshot by using the
4244.Nm zfs Cm destroy
4245command return
4246.Er EBUSY .
4247.Bl -tag -width "-r"
4248.It Fl r
4249Specifies that a hold with the given tag is applied recursively to the snapshots
4250of all descendent file systems.
4251.El
4252.It Xo
4253.Nm
4254.Cm holds
4255.Op Fl r
4256.Ar snapshot Ns ...
4257.Xc
4258Lists all existing user references for the given snapshot or snapshots.
4259.Bl -tag -width "-r"
4260.It Fl r
4261Lists the holds that are set on the named descendent snapshots, in addition to
4262listing the holds on the named snapshot.
4263.El
4264.It Xo
4265.Nm
4266.Cm release
4267.Op Fl r
4268.Ar tag Ar snapshot Ns ...
4269.Xc
4270Removes a single reference, named with the
4271.Ar tag
4272argument, from the specified snapshot or snapshots.
4273The tag must already exist for each snapshot.
4274If a hold exists on a snapshot, attempts to destroy that snapshot by using the
4275.Nm zfs Cm destroy
4276command return
4277.Er EBUSY .
4278.Bl -tag -width "-r"
4279.It Fl r
4280Recursively releases a hold with the given tag on the snapshots of all
4281descendent file systems.
4282.El
4283.It Xo
4284.Nm
4285.Cm diff
4286.Op Fl FHt
4287.Ar snapshot Ar snapshot Ns | Ns Ar filesystem
4288.Xc
4289Display the difference between a snapshot of a given filesystem and another
4290snapshot of that filesystem from a later time or the current contents of the
4291filesystem.
4292The first column is a character indicating the type of change, the other columns
4293indicate pathname, new pathname
4294.Pq in case of rename ,
4295change in link count, and optionally file type and/or change time.
4296The types of change are:
4297.Bd -literal
4298-       The path has been removed
4299+       The path has been created
4300M       The path has been modified
4301R       The path has been renamed
4302.Ed
4303.Bl -tag -width "-F"
4304.It Fl F
4305Display an indication of the type of file, in a manner similar to the
4306.Fl
4307option of
4308.Xr ls 1 .
4309.Bd -literal
4310B       Block device
4311C       Character device
4312/       Directory
4313>       Door
4314|       Named pipe
4315@       Symbolic link
4316P       Event port
4317=       Socket
4318F       Regular file
4319.Ed
4320.It Fl H
4321Give more parsable tab-separated output, without header lines and without
4322arrows.
4323.It Fl t
4324Display the path's inode change time as the first column of output.
4325.El
4326.It Xo
4327.Nm
4328.Cm program
4329.Op Fl jn
4330.Op Fl t Ar timeout
4331.Op Fl m Ar memory_limit
4332.Ar pool script
4333.Op Ar arg1 No ...
4334.Xc
4335Executes
4336.Ar script
4337as a ZFS channel program on
4338.Ar pool .
4339The ZFS channel
4340program interface allows ZFS administrative operations to be run
4341programmatically via a Lua script.
4342The entire script is executed atomically, with no other administrative
4343operations taking effect concurrently.
4344A library of ZFS calls is made available to channel program scripts.
4345Channel programs may only be run with root privileges.
4346.sp
4347For full documentation of the ZFS channel program interface, see the manual
4348page for
4349.Xr zfs-program 8 .
4350.Bl -tag -width ""
4351.It Fl j
4352Display channel program output in JSON format.
4353When this flag is specified and standard output is empty -
4354channel program encountered an error.
4355The details of such an error will be printed to standard error in plain text.
4356.It Fl n
4357Executes a read-only channel program, which runs faster.
4358The program cannot change on-disk state by calling functions from
4359the zfs.sync submodule.
4360The program can be used to gather information such as properties and
4361determining if changes would succeed (zfs.check.*).
4362Without this flag, all pending changes must be synced to disk before
4363a channel program can complete.
4364.It Fl t Ar timeout
4365Execution time limit, in milliseconds.
4366If a channel program executes for longer than the provided timeout, it will
4367be stopped and an error will be returned.
4368The default timeout is 1000 ms, and can be set to a maximum of 10000 ms.
4369.It Fl m Ar memory-limit
4370Memory limit, in bytes.
4371If a channel program attempts to allocate more memory than the given limit,
4372it will be stopped and an error returned.
4373The default memory limit is 10 MB, and can be set to a maximum of 100 MB.
4374.sp
4375All remaining argument strings are passed directly to the channel program as
4376arguments.
4377See
4378.Xr zfs-program 8
4379for more information.
4380.El
4381.It Xo
4382.Nm Cm load-key
4383.Op Fl nr
4384.Op Fl L Ar keylocation
4385.Fl a Ns | Ns filesystem
4386.Xc
4387Use
4388.Ar keylocation
4389instead of the
4390.Sy keylocation
4391property.
4392This will not change the value of the property on the dataset.
4393Note that if used with either
4394.Fl r
4395or
4396.Fl a
4397.Ar keylocation
4398may only be given as
4399.Sy prompt .
4400.Bl -tag -width Ds
4401.It Fl a
4402Loads the keys for all encryption roots in all imported pools.
4403.It Fl n
4404Do a dry-run
4405.Cm load-key .
4406This will cause zfs to simply check that the provided key is correct.
4407This command may be run even if the key is already loaded.
4408.It Fl r
4409Recursively loads the keys for the specified filesystem and all descendent
4410encryption roots.
4411.El
4412.It Xo
4413.Nm
4414.Cm unload-key
4415.Op Fl r
4416.Fl a Ns | Ns Ar filesystem
4417.Xc
4418Unloads a key from ZFS, removing the ability to access the dataset and all of
4419its children that inherit the
4420.Sy encryption
4421property.
4422This requires that the dataset is not currently open or mounted.
4423Once the key is unloaded the
4424.Sy keystatus
4425property will be set to
4426.Sy unavailable .
4427.Bl -tag -width Ds
4428.It Fl a
4429Unloads the keys for all encryption roots in all imported pools.
4430.It Fl r
4431Recursively unloads the keys for the specified filesystem and all descendent
4432encryption roots.
4433.El
4434.It Xo
4435.Nm
4436.Cm change-key
4437.Op Fl il
4438.Op Fl o Sy keylocation Ns = Ns Ar value
4439.Op Fl o Sy keyformat Ns = Ns Ar value
4440.Op Fl o Sy pbkdf2iters Ns = Ns Ar value
4441.Ar filesystem
4442.Xc
4443Allows a user to change the encryption key used to access a dataset.
4444This command requires that the existing key for the dataset is already loaded
4445into ZFS.
4446This command may also be used to change the
4447.Sy keylocation , keyformat ,
4448and
4449.Sy pbkdf2iters
4450properties as needed.
4451If the dataset was not previously an encryption root it will become one.
4452Alternatively, the
4453.Fl i
4454flag may be provided to cause an encryption root to inherit the
4455parent's key instead.
4456.Bl -tag -width Ds
4457.It Fl i
4458Indicates that ZFS should make
4459.Ar filesystem
4460inherit the key of its parent.
4461Note that this command can only be run on an encryption root that has an
4462encrypted parent.
4463.It Fl l
4464Ensures the key is loaded before attempting to change the key.
4465This is effectively equivalent to
4466.Qq Nm Cm load-key Ar filesystem ; Nm Cm change-key Ar filesystem .
4467.It Fl o Sy property Ns = Ns Ar value
4468Allows the user to set encryption key properties
4469.Pq
4470.Sy keyformat , keylocation ,
4471and
4472.Sy pbkdf2iters
4473while changing the key.
4474This is the only way to alter
4475.Sy keyformat
4476and
4477.Sy pbkdf2iters
4478after the dataset has been created.
4479.El
4480.El
4481.Sh EXIT STATUS
4482The
4483.Nm
4484utility exits 0 on success, 1 if an error occurs, and 2 if invalid command line
4485options were specified.
4486.Sh EXAMPLES
4487.Bl -tag -width ""
4488.It Sy Example 1 No Creating a ZFS File System Hierarchy
4489The following commands create a file system named
4490.Em pool/home
4491and a file system named
4492.Em pool/home/bob .
4493The mount point
4494.Pa /export/home
4495is set for the parent file system, and is automatically inherited by the child
4496file system.
4497.Bd -literal
4498# zfs create pool/home
4499# zfs set mountpoint=/export/home pool/home
4500# zfs create pool/home/bob
4501.Ed
4502.It Sy Example 2 No Creating a ZFS Snapshot
4503The following command creates a snapshot named
4504.Sy yesterday .
4505This snapshot is mounted on demand in the
4506.Pa .zfs/snapshot
4507directory at the root of the
4508.Em pool/home/bob
4509file system.
4510.Bd -literal
4511# zfs snapshot pool/home/bob@yesterday
4512.Ed
4513.It Sy Example 3 No Creating and Destroying Multiple Snapshots
4514The following command creates snapshots named
4515.Sy yesterday
4516of
4517.Em pool/home
4518and all of its descendent file systems.
4519Each snapshot is mounted on demand in the
4520.Pa .zfs/snapshot
4521directory at the root of its file system.
4522The second command destroys the newly created snapshots.
4523.Bd -literal
4524# zfs snapshot -r pool/home@yesterday
4525# zfs destroy -r pool/home@yesterday
4526.Ed
4527.It Sy Example 4 No Disabling and Enabling File System Compression
4528The following command disables the
4529.Sy compression
4530property for all file systems under
4531.Em pool/home .
4532The next command explicitly enables
4533.Sy compression
4534for
4535.Em pool/home/anne .
4536.Bd -literal
4537# zfs set compression=off pool/home
4538# zfs set compression=on pool/home/anne
4539.Ed
4540.It Sy Example 5 No Listing ZFS Datasets
4541The following command lists all active file systems and volumes in the system.
4542Snapshots are displayed if the
4543.Sy listsnaps
4544property is
4545.Sy on .
4546The default is
4547.Sy off .
4548See
4549.Xr zpool 8
4550for more information on pool properties.
4551.Bd -literal
4552# zfs list
4553NAME                      USED  AVAIL  REFER  MOUNTPOINT
4554pool                      450K   457G    18K  /pool
4555pool/home                 315K   457G    21K  /export/home
4556pool/home/anne             18K   457G    18K  /export/home/anne
4557pool/home/bob             276K   457G   276K  /export/home/bob
4558.Ed
4559.It Sy Example 6 No Setting a Quota on a ZFS File System
4560The following command sets a quota of 50 Gbytes for
4561.Em pool/home/bob .
4562.Bd -literal
4563# zfs set quota=50G pool/home/bob
4564.Ed
4565.It Sy Example 7 No Listing ZFS Properties
4566The following command lists all properties for
4567.Em pool/home/bob .
4568.Bd -literal
4569# zfs get all pool/home/bob
4570NAME           PROPERTY              VALUE                  SOURCE
4571pool/home/bob  type                  filesystem             -
4572pool/home/bob  creation              Tue Jul 21 15:53 2009  -
4573pool/home/bob  used                  21K                    -
4574pool/home/bob  available             20.0G                  -
4575pool/home/bob  referenced            21K                    -
4576pool/home/bob  compressratio         1.00x                  -
4577pool/home/bob  mounted               yes                    -
4578pool/home/bob  quota                 20G                    local
4579pool/home/bob  reservation           none                   default
4580pool/home/bob  recordsize            128K                   default
4581pool/home/bob  mountpoint            /pool/home/bob         default
4582pool/home/bob  sharenfs              off                    default
4583pool/home/bob  checksum              on                     default
4584pool/home/bob  compression           on                     local
4585pool/home/bob  atime                 on                     default
4586pool/home/bob  devices               on                     default
4587pool/home/bob  exec                  on                     default
4588pool/home/bob  setuid                on                     default
4589pool/home/bob  readonly              off                    default
4590pool/home/bob  zoned                 off                    default
4591pool/home/bob  snapdir               hidden                 default
4592pool/home/bob  aclmode               discard                default
4593pool/home/bob  aclinherit            restricted             default
4594pool/home/bob  canmount              on                     default
4595pool/home/bob  xattr                 on                     default
4596pool/home/bob  copies                1                      default
4597pool/home/bob  version               4                      -
4598pool/home/bob  utf8only              off                    -
4599pool/home/bob  normalization         none                   -
4600pool/home/bob  casesensitivity       sensitive              -
4601pool/home/bob  vscan                 off                    default
4602pool/home/bob  nbmand                off                    default
4603pool/home/bob  sharesmb              off                    default
4604pool/home/bob  refquota              none                   default
4605pool/home/bob  refreservation        none                   default
4606pool/home/bob  primarycache          all                    default
4607pool/home/bob  secondarycache        all                    default
4608pool/home/bob  usedbysnapshots       0                      -
4609pool/home/bob  usedbydataset         21K                    -
4610pool/home/bob  usedbychildren        0                      -
4611pool/home/bob  usedbyrefreservation  0                      -
4612.Ed
4613.Pp
4614The following command gets a single property value.
4615.Bd -literal
4616# zfs get -H -o value compression pool/home/bob
4617on
4618.Ed
4619The following command lists all properties with local settings for
4620.Em pool/home/bob .
4621.Bd -literal
4622# zfs get -r -s local -o name,property,value all pool/home/bob
4623NAME           PROPERTY              VALUE
4624pool/home/bob  quota                 20G
4625pool/home/bob  compression           on
4626.Ed
4627.It Sy Example 8 No Rolling Back a ZFS File System
4628The following command reverts the contents of
4629.Em pool/home/anne
4630to the snapshot named
4631.Sy yesterday ,
4632deleting all intermediate snapshots.
4633.Bd -literal
4634# zfs rollback -r pool/home/anne@yesterday
4635.Ed
4636.It Sy Example 9 No Creating a ZFS Clone
4637The following command creates a writable file system whose initial contents are
4638the same as
4639.Em pool/home/bob@yesterday .
4640.Bd -literal
4641# zfs clone pool/home/bob@yesterday pool/clone
4642.Ed
4643.It Sy Example 10 No Promoting a ZFS Clone
4644The following commands illustrate how to test out changes to a file system, and
4645then replace the original file system with the changed one, using clones, clone
4646promotion, and renaming:
4647.Bd -literal
4648# zfs create pool/project/production
4649  populate /pool/project/production with data
4650# zfs snapshot pool/project/production@today
4651# zfs clone pool/project/production@today pool/project/beta
4652  make changes to /pool/project/beta and test them
4653# zfs promote pool/project/beta
4654# zfs rename pool/project/production pool/project/legacy
4655# zfs rename pool/project/beta pool/project/production
4656  once the legacy version is no longer needed, it can be destroyed
4657# zfs destroy pool/project/legacy
4658.Ed
4659.It Sy Example 11 No Inheriting ZFS Properties
4660The following command causes
4661.Em pool/home/bob
4662and
4663.Em pool/home/anne
4664to inherit the
4665.Sy checksum
4666property from their parent.
4667.Bd -literal
4668# zfs inherit checksum pool/home/bob pool/home/anne
4669.Ed
4670.It Sy Example 12 No Remotely Replicating ZFS Data
4671The following commands send a full stream and then an incremental stream to a
4672remote machine, restoring them into
4673.Em poolB/received/fs@a
4674and
4675.Em poolB/received/fs@b ,
4676respectively.
4677.Em poolB
4678must contain the file system
4679.Em poolB/received ,
4680and must not initially contain
4681.Em poolB/received/fs .
4682.Bd -literal
4683# zfs send pool/fs@a | \e
4684  ssh host zfs receive poolB/received/fs@a
4685# zfs send -i a pool/fs@b | \e
4686  ssh host zfs receive poolB/received/fs
4687.Ed
4688.It Sy Example 13 No Using the zfs receive -d Option
4689The following command sends a full stream of
4690.Em poolA/fsA/fsB@snap
4691to a remote machine, receiving it into
4692.Em poolB/received/fsA/fsB@snap .
4693The
4694.Em fsA/fsB@snap
4695portion of the received snapshot's name is determined from the name of the sent
4696snapshot.
4697.Em poolB
4698must contain the file system
4699.Em poolB/received .
4700If
4701.Em poolB/received/fsA
4702does not exist, it is created as an empty file system.
4703.Bd -literal
4704# zfs send poolA/fsA/fsB@snap | \e
4705  ssh host zfs receive -d poolB/received
4706.Ed
4707.It Sy Example 14 No Setting User Properties
4708The following example sets the user-defined
4709.Sy com.example:department
4710property for a dataset.
4711.Bd -literal
4712# zfs set com.example:department=12345 tank/accounting
4713.Ed
4714.It Sy Example 15 No Performing a Rolling Snapshot
4715The following example shows how to maintain a history of snapshots with a
4716consistent naming scheme.
4717To keep a week's worth of snapshots, the user destroys the oldest snapshot,
4718renames the remaining snapshots, and then creates a new snapshot, as follows:
4719.Bd -literal
4720# zfs destroy -r pool/users@7daysago
4721# zfs rename -r pool/users@6daysago @7daysago
4722# zfs rename -r pool/users@5daysago @6daysago
4723# zfs rename -r pool/users@4daysago @5daysago
4724# zfs rename -r pool/users@3daysago @4daysago
4725# zfs rename -r pool/users@2daysago @3daysago
4726# zfs rename -r pool/users@yesterday @2daysago
4727# zfs rename -r pool/users@today @yesterday
4728# zfs snapshot -r pool/users@today
4729.Ed
4730.It Sy Example 16 No Setting sharenfs Property Options on a ZFS File System
4731The following commands show how to set
4732.Sy sharenfs
4733property options to enable
4734.Sy rw
4735access for a set of
4736.Sy IP
4737addresses and to enable root access for system
4738.Sy neo
4739on the
4740.Em tank/home
4741file system.
4742.Bd -literal
4743# zfs set sharenfs='rw=@123.123.0.0/16,root=neo' tank/home
4744.Ed
4745.Pp
4746If you are using
4747.Sy DNS
4748for host name resolution, specify the fully qualified hostname.
4749.It Sy Example 17 No Delegating ZFS Administration Permissions on a ZFS Dataset
4750The following example shows how to set permissions so that user
4751.Sy cindys
4752can create, destroy, mount, and take snapshots on
4753.Em tank/cindys .
4754The permissions on
4755.Em tank/cindys
4756are also displayed.
4757.Bd -literal
4758# zfs allow cindys create,destroy,mount,snapshot tank/cindys
4759# zfs allow tank/cindys
4760---- Permissions on tank/cindys --------------------------------------
4761Local+Descendent permissions:
4762        user cindys create,destroy,mount,snapshot
4763.Ed
4764.Pp
4765Because the
4766.Em tank/cindys
4767mount point permission is set to 755 by default, user
4768.Sy cindys
4769will be unable to mount file systems under
4770.Em tank/cindys .
4771Add an ACE similar to the following syntax to provide mount point access:
4772.Bd -literal
4773# chmod A+user:cindys:add_subdirectory:allow /tank/cindys
4774.Ed
4775.It Sy Example 18 No Delegating Create Time Permissions on a ZFS Dataset
4776The following example shows how to grant anyone in the group
4777.Sy staff
4778to create file systems in
4779.Em tank/users .
4780This syntax also allows staff members to destroy their own file systems, but not
4781destroy anyone else's file system.
4782The permissions on
4783.Em tank/users
4784are also displayed.
4785.Bd -literal
4786# zfs allow staff create,mount tank/users
4787# zfs allow -c destroy tank/users
4788# zfs allow tank/users
4789---- Permissions on tank/users ---------------------------------------
4790Permission sets:
4791        destroy
4792Local+Descendent permissions:
4793        group staff create,mount
4794.Ed
4795.It Sy Example 19 No Defining and Granting a Permission Set on a ZFS Dataset
4796The following example shows how to define and grant a permission set on the
4797.Em tank/users
4798file system.
4799The permissions on
4800.Em tank/users
4801are also displayed.
4802.Bd -literal
4803# zfs allow -s @pset create,destroy,snapshot,mount tank/users
4804# zfs allow staff @pset tank/users
4805# zfs allow tank/users
4806---- Permissions on tank/users ---------------------------------------
4807Permission sets:
4808        @pset create,destroy,mount,snapshot
4809Local+Descendent permissions:
4810        group staff @pset
4811.Ed
4812.It Sy Example 20 No Delegating Property Permissions on a ZFS Dataset
4813The following example shows to grant the ability to set quotas and reservations
4814on the
4815.Em users/home
4816file system.
4817The permissions on
4818.Em users/home
4819are also displayed.
4820.Bd -literal
4821# zfs allow cindys quota,reservation users/home
4822# zfs allow users/home
4823---- Permissions on users/home ---------------------------------------
4824Local+Descendent permissions:
4825        user cindys quota,reservation
4826cindys% zfs set quota=10G users/home/marks
4827cindys% zfs get quota users/home/marks
4828NAME              PROPERTY  VALUE  SOURCE
4829users/home/marks  quota     10G    local
4830.Ed
4831.It Sy Example 21 No Removing ZFS Delegated Permissions on a ZFS Dataset
4832The following example shows how to remove the snapshot permission from the
4833.Sy staff
4834group on the
4835.Em tank/users
4836file system.
4837The permissions on
4838.Em tank/users
4839are also displayed.
4840.Bd -literal
4841# zfs unallow staff snapshot tank/users
4842# zfs allow tank/users
4843---- Permissions on tank/users ---------------------------------------
4844Permission sets:
4845        @pset create,destroy,mount,snapshot
4846Local+Descendent permissions:
4847        group staff @pset
4848.Ed
4849.It Sy Example 22 No Showing the differences between a snapshot and a ZFS Dataset
4850The following example shows how to see what has changed between a prior
4851snapshot of a ZFS dataset and its current state.
4852The
4853.Fl F
4854option is used to indicate type information for the files affected.
4855.Bd -literal
4856# zfs diff -F tank/test@before tank/test
4857M       /       /tank/test/
4858M       F       /tank/test/linked      (+1)
4859R       F       /tank/test/oldname -> /tank/test/newname
4860-       F       /tank/test/deleted
4861+       F       /tank/test/created
4862M       F       /tank/test/modified
4863.Ed
4864.El
4865.Sh INTERFACE STABILITY
4866.Sy Committed .
4867.Sh SEE ALSO
4868.Xr gzip 1 ,
4869.Xr ssh 1 ,
4870.Xr chmod 2 ,
4871.Xr stat 2 ,
4872.Xr write 2 ,
4873.Xr fsync 3C ,
4874.Xr dfstab 5 ,
4875.Xr acl 7 ,
4876.Xr attributes 7 ,
4877.Xr mount 8 ,
4878.Xr share 8 ,
4879.Xr sharemgr 8 ,
4880.Xr unshare 8 ,
4881.Xr zfs-program 8 ,
4882.Xr zonecfg 8 ,
4883.Xr zpool 8
4884