6 btrfs-quota - control the global quota status of a btrfs filesystem
10 *btrfs quota* <subcommand> <args>
14 The commands under *btrfs quota* are used to affect the global status of quotas
15 of a btrfs filesystem. The quota groups (qgroups) are managed by the subcommand
18 NOTE: the qgroups are different than the traditional user quotas and designed
19 to track shared and exclusive data per-subvolume. Please refer to the section
20 'HIERARCHICAL QUOTA GROUP CONCEPTS' for a detailed description.
22 PERFORMANCE IMPLICATIONS
23 ~~~~~~~~~~~~~~~~~~~~~~~~
25 When the quotas are turned on, they affect all extent processing, taking a
26 performance hit. It is not recommended to turn on qgroups unless the user
27 intends to actually use them.
32 The qgroup implementation has turned out to be quite difficult as it affects
33 the core of the filesystem operation. The users have hit various corner cases
34 over time, eg. wrong accounting or system instability. The situation is
35 gradually improving but currently (4.7) there are still issues found and fixed.
37 HIERARCHICAL QUOTA GROUP CONCEPTS
38 ---------------------------------
40 The concept of quota has a long-standing tradition in the Unix world. Ever
41 since computers allow multiple users to work simultaneously in one filesystem,
42 there is the need to prevent one user from using up the entire space. Every
43 user should get his fair share of the available resources.
45 In case of files, the solution is quite straightforward. Each file has an
46 'owner' recorded along with it, and it has a size. Traditional quota just
47 restricts the total size of all files that are owned by a user. The concept is
48 quite flexible: if a user hits his quota limit, the administrator can raise it
51 On the other hand, the traditional approach has only a poor solution to
53 At installation time, the harddisk can be partitioned so that every directory
54 (eg. /usr, /var/, ...) that needs a limit gets its own partition. The obvious
55 problem is, that those limits cannot be changed without a reinstallation. The
56 btrfs subvolume feature builds a bridge. Subvolumes correspond in many ways to
57 partitions, as every subvolume looks like its own filesystem. With subvolume
58 quota, it is now possible to restrict each subvolume like a partition, but keep
59 the flexibility of quota. The space for each subvolume can be expanded or
60 restricted on the fly.
62 As subvolumes are the basis for snapshots, interesting questions arise as to
63 how to account used space in the presence of snapshots. If you have a file
64 shared between a subvolume and a snapshot, whom to account the file to? The
65 creator? Both? What if the file gets modified in the snapshot, should only
66 these changes be accounted to it? But wait, both the snapshot and the subvolume
67 belong to the same user home. I just want to limit the total space used by
68 both! But somebody else might not want to charge the snapshots to the users.
70 Btrfs subvolume quota solves these problems by introducing groups of subvolumes
71 and let the user put limits on them. It is even possible to have groups of
72 groups. In the following, we refer to them as 'qgroups'.
74 Each qgroup primarily tracks two numbers, the amount of total referenced
75 space and the amount of exclusively referenced space.
78 space is the amount of data that can be reached from any of the subvolumes
79 contained in the qgroup, while
81 is the amount of data where all references to this data can be reached
82 from within this qgroup.
84 SUBVOLUME QUOTA GROUPS
85 ~~~~~~~~~~~~~~~~~~~~~~
87 The basic notion of the Subvolume Quota feature is the qouta group, short
88 qgroup. Qgroups are notated as 'level/id', eg. the qgroup 3/2 is a qgroup of
89 level 3. For level 0, the leading '0/' can be omitted.
90 Qgroups of level 0 get created automatically when a subvolume/snapshot gets
91 created. The ID of the qgroup corresponds to the ID of the subvolume, so 0/5
92 is the qgroup for the root subvolume.
93 For the *btrfs qgroup* command, the path to the subvolume can also be used
94 instead of '0/ID'. For all higher levels, the ID can be chosen freely.
96 Each qgroup can contain a set of lower level qgroups, thus creating a hierarchy
97 of qgroups. Figure 1 shows an example qgroup tree.
99 // TODO: insert Figure 1
101 At the bottom, some extents are depicted showing which qgroups reference which
102 extents. It is important to understand the notion of 'referenced' vs
103 'exclusive'. In the example, qgroup 0/2 references extents 2 and 3, while 1/2
104 references extents 2-4, 2/1 references all extents.
106 On the other hand, extent 1 is exclusive to 0/1, extent 2 is exclusive to 0/2,
107 while extent 3 is neither exclusive to 0/2 nor to 0/3. But because both
108 references can be reached from 1/2, extent 3 is exclusive to 1/2. All extents
109 are exclusive to 2/1.
111 So exclusive does not mean there is no other way to reach the extent, but it
112 does mean that if you delete all subvolumes contained in a qgroup, the extent
115 Exclusive of a qgroup conveys the useful information how much space will be
116 freed in case all subvolumes of the qgroup get deleted.
118 All data extents are accounted this way. Metadata that belongs to a specific
119 subvolume (i.e. its filesystem tree) is also accounted. Checksums and extent
120 allocation information are not accounted.
122 In turn, the referenced count of a qgroup can be limited. All writes beyond
123 this limit will lead to a 'Quota Exceeded' error.
128 Things get a bit more complicated when new subvolumes or snapshots are created.
129 The case of (empty) subvolumes is still quite easy. If a subvolume should be
130 part of a qgroup, it has to be added to the qgroup at creation time. To add it
131 at a later time, it would be necessary to at least rescan the full subvolume
132 for a proper accounting.
134 Creation of a snapshot is the hard case. Obviously, the snapshot will
135 reference the exact amount of space as its source, and both source and
136 destination now have an exclusive count of 0 (the filesystem nodesize to be
137 precise, as the roots of the trees are not shared). But what about qgroups of
138 higher levels? If the qgroup contains both the source and the destination,
139 nothing changes. If the qgroup contains only the source, it might lose some
142 But how much? The tempting answer is, subtract all exclusive of the source from
143 the qgroup, but that is wrong, or at least not enough. There could have been
144 an extent that is referenced from the source and another subvolume from that
145 qgroup. This extent would have been exclusive to the qgroup, but not to the
146 source subvolume. With the creation of the snapshot, the qgroup would also
147 lose this extent from its exclusive set.
149 So how can this problem be solved? In the instant the snapshot gets created, we
150 already have to know the correct exclusive count. We need to have a second
151 qgroup that contains all the subvolumes as the first qgroup, except the
152 subvolume we want to snapshot. The moment we create the snapshot, the
153 exclusive count from the second qgroup needs to be copied to the first qgroup,
154 as it represents the correct value. The second qgroup is called a tracking
155 qgroup. It is only there in case a snapshot is needed.
160 Below are some usecases that do not mean to be extensive. You can find your
161 own way how to integrate qgroups.
163 ==== SINGLE-USER MACHINE ====
165 `Replacement for partitions`
167 The simplest use case is to use qgroups as simple replacement for partitions.
168 Btrfs takes the disk as a whole, and /, /usr, /var etc. are created as
169 subvolumes. As each subvolume gets it own qgroup automatically, they can
170 simply be restricted. No hierarchy is needed for that.
172 `Track usage of snapshots`
174 When a snapshot is taken, a qgroup for it will automatically be created with
175 the correct values. 'Referenced' will show how much is in it, possibly shared
176 with other subvolumes. 'Exclusive' will be the amount of space that gets freed
177 when the subvolume is deleted.
179 ==== MULTI-USER MACHINE ====
183 When you have several users on a machine, with home directories probably under
184 /home, you might want to restrict /home as a whole, while restricting every
185 user to an indiviual limit as well. This is easily accomplished by creating a
186 qgroup for /home , eg. 1/1, and assigning all user subvolumes to it.
187 Restricting this qgroup will limit /home, while every user subvolume can get
188 its own (lower) limit.
190 `Accounting snapshots to the user`
192 Let's say the user is allowed to create snapshots via some mechanism. It would
193 only be fair to account space used by the snapshots to the user. This does not
194 mean the user doubles his usage as soon as he takes a snapshot. Of course,
195 files that are present in his home and the snapshot should only be accounted
196 once. This can be accomplished by creating a qgroup for each user, say
197 '1/UID'. The user home and all snapshots are assigned to this qgroup.
198 Limiting it will extend the limit to all snapshots, counting files only once.
199 To limit /home as a whole, a higher level group 2/1 replacing 1/1 from the
200 previous example is needed, with all user qgroups assigned to it.
202 `Do not account snapshots`
204 On the other hand, when the snapshots get created automatically, the user has
205 no chance to control them, so the space used by them should not be accounted to
206 him. This is already the case when creating snapshots in the example from
207 the previous section.
209 `Snapshots for backup purposes`
211 This scenario is a mixture of the previous two. The user can create snapshots,
212 but some snapshots for backup purposes are being created by the system. The
213 user's snapshots should be accounted to the user, not the system. The solution
214 is similar to the one from section 'Accounting snapshots to the user', but do
215 not assign system snapshots to user's qgroup.
220 Disable subvolume quota support for a filesystem.
223 Enable subvolume quota support for a filesystem.
225 *rescan* [-s] <path>::
226 Trash all qgroup numbers and scan the metadata again with the current config.
231 show status of a running rescan operation.
233 wait for rescan operation to finish(can be already in progress).
237 *btrfs quota* returns a zero exit status if it succeeds. Non zero is
238 returned in case of failure.
242 *btrfs* is part of btrfs-progs.
243 Please refer to the btrfs wiki http://btrfs.wiki.kernel.org for
249 `btrfs-subvolume`(8),