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.\" Copyright (c) 2009 Sun Microsystems, Inc. All Rights Reserved. 22.\" Copyright 2011 Joshua M. Clulow <josh@sysmgr.org> 23.\" Copyright (c) 2011, 2019 by Delphix. All rights reserved. 24.\" Copyright (c) 2013 by Saso Kiselkov. All rights reserved. 25.\" Copyright (c) 2014, Joyent, Inc. All rights reserved. 26.\" Copyright (c) 2014 by Adam Stevko. All rights reserved. 27.\" Copyright (c) 2014 Integros [integros.com] 28.\" Copyright 2019 Richard Laager. All rights reserved. 29.\" Copyright 2018 Nexenta Systems, Inc. 30.\" Copyright 2019 Joyent, Inc. 31.\" 32.Dd June 30, 2019 33.Dt ZFSCONCEPTS 7 34.Os 35. 36.Sh NAME 37.Nm zfsconcepts 38.Nd overview of ZFS concepts 39. 40.Sh DESCRIPTION 41.Ss ZFS File System Hierarchy 42A ZFS storage pool is a logical collection of devices that provide space for 43datasets. 44A storage pool is also the root of the ZFS file system hierarchy. 45.Pp 46The root of the pool can be accessed as a file system, such as mounting and 47unmounting, taking snapshots, and setting properties. 48The physical storage characteristics, however, are managed by the 49.Xr zpool 8 50command. 51.Pp 52See 53.Xr zpool 8 54for more information on creating and administering pools. 55.Ss Snapshots 56A snapshot is a read-only copy of a file system or volume. 57Snapshots can be created extremely quickly, and initially consume no additional 58space within the pool. 59As data within the active dataset changes, the snapshot consumes more data than 60would otherwise be shared with the active dataset. 61.Pp 62Snapshots can have arbitrary names. 63Snapshots of volumes can be cloned or rolled back, visibility is determined 64by the 65.Sy snapdev 66property of the parent volume. 67.Pp 68File system snapshots can be accessed under the 69.Pa .zfs/snapshot 70directory in the root of the file system. 71Snapshots are automatically mounted on demand and may be unmounted at regular 72intervals. 73The visibility of the 74.Pa .zfs 75directory can be controlled by the 76.Sy snapdir 77property. 78.Ss Bookmarks 79A bookmark is like a snapshot, a read-only copy of a file system or volume. 80Bookmarks can be created extremely quickly, compared to snapshots, and they 81consume no additional space within the pool. 82Bookmarks can also have arbitrary names, much like snapshots. 83.Pp 84Unlike snapshots, bookmarks can not be accessed through the filesystem in any way. 85From a storage standpoint a bookmark just provides a way to reference 86when a snapshot was created as a distinct object. 87Bookmarks are initially tied to a snapshot, not the filesystem or volume, 88and they will survive if the snapshot itself is destroyed. 89Since they are very light weight there's little incentive to destroy them. 90.Ss Clones 91A clone is a writable volume or file system whose initial contents are the same 92as another dataset. 93As with snapshots, creating a clone is nearly instantaneous, and initially 94consumes no additional space. 95.Pp 96Clones can only be created from a snapshot. 97When a snapshot is cloned, it creates an implicit dependency between the parent 98and child. 99Even though the clone is created somewhere else in the dataset hierarchy, the 100original snapshot cannot be destroyed as long as a clone exists. 101The 102.Sy origin 103property exposes this dependency, and the 104.Cm destroy 105command lists any such dependencies, if they exist. 106.Pp 107The clone parent-child dependency relationship can be reversed by using the 108.Cm promote 109subcommand. 110This causes the 111.Qq origin 112file system to become a clone of the specified file system, which makes it 113possible to destroy the file system that the clone was created from. 114.Ss "Mount Points" 115Creating a ZFS file system is a simple operation, so the number of file systems 116per system is likely to be numerous. 117To cope with this, ZFS automatically manages mounting and unmounting file 118systems without the need to edit the 119.Pa /etc/fstab 120file. 121All automatically managed file systems are mounted by ZFS at boot time. 122.Pp 123By default, file systems are mounted under 124.Pa /path , 125where 126.Ar path 127is the name of the file system in the ZFS namespace. 128Directories are created and destroyed as needed. 129.Pp 130A file system can also have a mount point set in the 131.Sy mountpoint 132property. 133This directory is created as needed, and ZFS automatically mounts the file 134system when the 135.Nm zfs Cm mount Fl a 136command is invoked 137.Po without editing 138.Pa /etc/fstab 139.Pc . 140The 141.Sy mountpoint 142property can be inherited, so if 143.Em pool/home 144has a mount point of 145.Pa /export/stuff , 146then 147.Em pool/home/user 148automatically inherits a mount point of 149.Pa /export/stuff/user . 150.Pp 151A file system 152.Sy mountpoint 153property of 154.Sy none 155prevents the file system from being mounted. 156.Pp 157If needed, ZFS file systems can also be managed with traditional tools 158.Po 159.Nm mount , 160.Nm umount , 161.Pa /etc/fstab 162.Pc . 163If a file system's mount point is set to 164.Sy legacy , 165ZFS makes no attempt to manage the file system, and the administrator is 166responsible for mounting and unmounting the file system. 167Because pools must 168be imported before a legacy mount can succeed, administrators should ensure 169that legacy mounts are only attempted after the zpool import process 170finishes at boot time. 171For example, on machines using systemd, the mount option 172.Pp 173.Nm x-systemd.requires=zfs-import.target 174.Pp 175will ensure that the zfs-import completes before systemd attempts mounting 176the filesystem. 177See 178.Xr systemd.mount 5 179for details. 180.Ss Deduplication 181Deduplication is the process for removing redundant data at the block level, 182reducing the total amount of data stored. 183If a file system has the 184.Sy dedup 185property enabled, duplicate data blocks are removed synchronously. 186The result 187is that only unique data is stored and common components are shared among files. 188.Pp 189Deduplicating data is a very resource-intensive operation. 190It is generally recommended that you have at least 1.25 GiB of RAM 191per 1 TiB of storage when you enable deduplication. 192Calculating the exact requirement depends heavily 193on the type of data stored in the pool. 194.Pp 195Enabling deduplication on an improperly-designed system can result in 196performance issues (slow I/O and administrative operations). 197It can potentially lead to problems importing a pool due to memory exhaustion. 198Deduplication can consume significant processing power (CPU) and memory as well 199as generate additional disk I/O. 200.Pp 201Before creating a pool with deduplication enabled, ensure that you have planned 202your hardware requirements appropriately and implemented appropriate recovery 203practices, such as regular backups. 204Consider using the 205.Sy compression 206property as a less resource-intensive alternative. 207