path: root/upstream/debian-bookworm/man8/btrfs-balance.8

.\" Man page generated from reStructuredText.
.
.
.nr rst2man-indent-level 0
.
.de1 rstReportMargin
\\$1 \\n[an-margin]
level \\n[rst2man-indent-level]
level margin: \\n[rst2man-indent\\n[rst2man-indent-level]]
-
\\n[rst2man-indent0]
\\n[rst2man-indent1]
\\n[rst2man-indent2]
..
.de1 INDENT
.\" .rstReportMargin pre:
. RS \\$1
. nr rst2man-indent\\n[rst2man-indent-level] \\n[an-margin]
. nr rst2man-indent-level +1
.\" .rstReportMargin post:
..
.de UNINDENT
. RE
.\" indent \\n[an-margin]
.\" old: \\n[rst2man-indent\\n[rst2man-indent-level]]
.nr rst2man-indent-level -1
.\" new: \\n[rst2man-indent\\n[rst2man-indent-level]]
.in \\n[rst2man-indent\\n[rst2man-indent-level]]u
..
.TH "BTRFS-BALANCE" "8" "Feb 28, 2023" "6.2" "BTRFS"
.SH NAME
btrfs-balance \- balance block groups on a btrfs filesystem
.SH SYNOPSIS
.sp
\fBbtrfs balance\fP <subcommand> <args>
.SH DESCRIPTION
.sp
The primary purpose of the balance feature is to spread block groups across
all devices so they match constraints defined by the respective profiles. See
\fI\%mkfs.btrfs(8)\fP section \fIPROFILES\fP for more details.
The scope of the balancing process can be further tuned by use of filters that
can select the block groups to process. Balance works only on a mounted
filesystem.  Extent sharing is preserved and reflinks are not broken.
Files are not defragmented nor recompressed, file extents are preserved
but the physical location on devices will change.
.sp
The balance operation is cancellable by the user. The on\-disk state of the
filesystem is always consistent so an unexpected interruption (e.g. system crash,
reboot) does not corrupt the filesystem. The progress of the balance operation
is temporarily stored as an internal state and will be resumed upon mount,
unless the mount option \fIskip_balance\fP is specified.
.sp
\fBWARNING:\fP
.INDENT 0.0
.INDENT 3.5
Running balance without filters will take a lot of time as it basically move
data/metadata from the whole filesystem and needs to update all block
pointers.
.UNINDENT
.UNINDENT
.sp
The filters can be used to perform following actions:
.INDENT 0.0
.IP \(bu 2
convert block group profiles (filter \fIconvert\fP)
.IP \(bu 2
make block group usage more compact  (filter \fIusage\fP)
.IP \(bu 2
perform actions only on a given device (filters \fIdevid\fP, \fIdrange\fP)
.UNINDENT
.sp
The filters can be applied to a combination of block group types (data,
metadata, system). Note that changing only the \fIsystem\fP type needs the force
option. Otherwise \fIsystem\fP gets automatically converted whenever \fImetadata\fP
profile is converted.
.sp
When metadata redundancy is reduced (e.g. from RAID1 to single) the force option
is also required and it is noted in system log.
.sp
\fBNOTE:\fP
.INDENT 0.0
.INDENT 3.5
The balance operation needs enough work space, i.e. space that is completely
unused in the filesystem, otherwise this may lead to ENOSPC reports.  See
the section \fIENOSPC\fP for more details.
.UNINDENT
.UNINDENT
.SH COMPATIBILITY
.sp
\fBNOTE:\fP
.INDENT 0.0
.INDENT 3.5
The balance subcommand also exists under the \fBbtrfs filesystem\fP namespace.
This still works for backward compatibility but is deprecated and should not
be used any more.
.UNINDENT
.UNINDENT
.sp
\fBNOTE:\fP
.INDENT 0.0
.INDENT 3.5
A short syntax \fBbtrfs balance <path>\fP works due to backward compatibility
but is deprecated and should not be used any more. Use \fBbtrfs balance start\fP
command instead.
.UNINDENT
.UNINDENT
.SH PERFORMANCE IMPLICATIONS
.sp
Balancing operations are very IO intensive and can also be quite CPU intensive,
impacting other ongoing filesystem operations. Typically large amounts of data
are copied from one location to another, with corresponding metadata updates.
.sp
Depending upon the block group layout, it can also be seek heavy. Performance
on rotational devices is noticeably worse compared to SSDs or fast arrays.
.SH SUBCOMMAND
.INDENT 0.0
.TP
.B cancel <path>
cancels a running or paused balance, the command will block and wait until the
current block group being processed completes
.sp
Since kernel 5.7 the response time of the cancellation is significantly
improved, on older kernels it might take a long time until currently
processed chunk is completely finished.
.TP
.B pause <path>
pause running balance operation, this will store the state of the balance
progress and used filters to the filesystem
.TP
.B resume <path>
resume interrupted balance, the balance status must be stored on the filesystem
from previous run, e.g. after it was paused or forcibly interrupted and mounted
again with \fIskip_balance\fP
.TP
.B start [options] <path>
start the balance operation according to the specified filters, without any filters
the data and metadata from the whole filesystem are moved. The process runs in
the foreground.
.sp
\fBNOTE:\fP
.INDENT 7.0
.INDENT 3.5
The balance command without filters will basically move everything in the
filesystem to a new physical location on devices (i.e. it does not affect the
logical properties of file extents like offsets within files and extent
sharing).  The run time is potentially very long, depending on the filesystem
size. To prevent starting a full balance by accident, the user is warned and
has a few seconds to cancel the operation before it starts.  The warning and
delay can be skipped with \fI\-\-full\-balance\fP option.
.UNINDENT
.UNINDENT
.sp
Please note that the filters must be written together with the \fI\-d\fP, \fI\-m\fP and
\fI\-s\fP options, because they\(aqre optional and bare \fI\-d\fP and \fI\-m\fP also work and
mean no filters.
.sp
\fBNOTE:\fP
.INDENT 7.0
.INDENT 3.5
When the target profile for conversion filter is \fIraid5\fP or \fIraid6\fP,
there\(aqs a safety timeout of 10 seconds to warn users about the status of the feature
.UNINDENT
.UNINDENT
.sp
\fBOptions\fP
.INDENT 7.0
.TP
.B \-d[<filters>]
act on data block groups, see \fIFILTERS\fP section for details about \fIfilters\fP
.TP
.B \-m[<filters>]
act on metadata chunks, see \fIFILTERS\fP section for details about \fIfilters\fP
.TP
.B \-s[<filters>]
act on system chunks (requires \fI\-f\fP), see \fIFILTERS\fP section for details about \fIfilters\fP\&.
.UNINDENT
.INDENT 7.0
.TP
.B  \-f
force a reduction of metadata integrity, e.g. when going from \fIraid1\fP to
\fIsingle\fP, or skip safety timeout when the target conversion profile is \fIraid5\fP
or \fIraid6\fP
.UNINDENT
.INDENT 7.0
.TP
.B \-\-background|\-\-bg
run the balance operation asynchronously in the background, uses \fBfork(2)\fP to
start the process that calls the kernel ioctl
.UNINDENT
.INDENT 7.0
.TP
.B  \-\-enqueue
wait if there\(aqs another exclusive operation running, otherwise continue
.TP
.B  \-v
(deprecated) alias for global \(aq\-v\(aq option
.UNINDENT
.TP
.B status [\-v] <path>
Show status of running or paused balance.
.sp
\fBOptions\fP
.INDENT 7.0
.TP
.B  \-v
(deprecated) alias for global \fI\-v\fP option
.UNINDENT
.UNINDENT
.SH FILTERS
.sp
From kernel 3.3 onwards, btrfs balance can limit its action to a subset of the
whole filesystem, and can be used to change the replication configuration (e.g.
moving data from single to RAID1). This functionality is accessed through the
\fI\-d\fP, \fI\-m\fP or \fI\-s\fP options to btrfs balance start, which filter on data,
metadata and system blocks respectively.
.sp
A filter has the following structure: \fItype[=params][,type=...]\fP
.sp
The available types are:
.INDENT 0.0
.TP
.B profiles=<profiles>
Balances only block groups with the given profiles. Parameters
are a list of profile names separated by \(dq\fI|\fP\(dq (pipe).
.TP
.B usage=<percent>, usage=<range>
Balances only block groups with usage under the given percentage. The
value of 0 is allowed and will clean up completely unused block groups, this
should not require any new work space allocated. You may want to use \fIusage=0\fP
in case balance is returning ENOSPC and your filesystem is not too full.
.sp
The argument may be a single value or a range. The single value \fIN\fP means \fIat
most N percent used\fP, equivalent to \fI\&..N\fP range syntax. Kernels prior to 4.4
accept only the single value format.
The minimum range boundary is inclusive, maximum is exclusive.
.TP
.B devid=<id>
Balances only block groups which have at least one chunk on the given
device. To list devices with ids use \fBbtrfs filesystem show\fP\&.
.TP
.B drange=<range>
Balance only block groups which overlap with the given byte range on any
device. Use in conjunction with \fIdevid\fP to filter on a specific device. The
parameter is a range specified as \fIstart..end\fP\&.
.TP
.B vrange=<range>
Balance only block groups which overlap with the given byte range in the
filesystem\(aqs internal virtual address space. This is the address space that
most reports from btrfs in the kernel log use. The parameter is a range
specified as \fIstart..end\fP\&.
.TP
.B convert=<profile>
Convert each selected block group to the given profile name identified by
parameters.
.sp
\fBNOTE:\fP
.INDENT 7.0
.INDENT 3.5
Starting with kernel 4.5, the \fIdata\fP chunks can be converted to/from the
\fIDUP\fP profile on a single device.
.UNINDENT
.UNINDENT
.sp
\fBNOTE:\fP
.INDENT 7.0
.INDENT 3.5
Starting with kernel 4.6, all profiles can be converted to/from \fIDUP\fP on
multi\-device filesystems.
.UNINDENT
.UNINDENT
.TP
.B limit=<number>, limit=<range>
Process only given number of chunks, after all filters are applied. This can be
used to specifically target a chunk in connection with other filters (\fIdrange\fP,
\fIvrange\fP) or just simply limit the amount of work done by a single balance run.
.sp
The argument may be a single value or a range. The single value \fIN\fP means \fIat
most N chunks\fP, equivalent to \fI\&..N\fP range syntax. Kernels prior to 4.4 accept
only the single value format.  The range minimum and maximum are inclusive.
.TP
.B stripes=<range>
Balance only block groups which have the given number of stripes. The parameter
is a range specified as \fIstart..end\fP\&. Makes sense for block group profiles that
utilize striping, i.e. RAID0/10/5/6.  The range minimum and maximum are
inclusive.
.TP
.B soft
Takes no parameters. Only has meaning when converting between profiles.
When doing convert from one profile to another and soft mode is on,
chunks that already have the target profile are left untouched.
This is useful e.g. when half of the filesystem was converted earlier but got
cancelled.
.sp
The soft mode switch is (like every other filter) per\-type.
For example, this means that we can convert metadata chunks the \(dqhard\(dq way
while converting data chunks selectively with soft switch.
.UNINDENT
.sp
Profile names, used in \fIprofiles\fP and \fIconvert\fP are one of: \fIraid0\fP, \fIraid1\fP,
\fIraid1c3\fP, \fIraid1c4\fP, \fIraid10\fP, \fIraid5\fP, \fIraid6\fP, \fIdup\fP, \fIsingle\fP\&.  The mixed
data/metadata profiles can be converted in the same way, but it\(aqs conversion
between mixed and non\-mixed is not implemented. For the constraints of the
profiles please refer to \fI\%mkfs.btrfs(8)\fP, section \fIPROFILES\fP\&.
.SH ENOSPC
.sp
The way balance operates, it usually needs to temporarily create a new block
group and move the old data there, before the old block group can be removed.
For that it needs the work space, otherwise it fails for ENOSPC reasons.
This is not the same ENOSPC as if the free space is exhausted. This refers to
the space on the level of block groups, which are bigger parts of the filesystem
that contain many file extents.
.sp
The free work space can be calculated from the output of the \fBbtrfs filesystem show\fP
command:
.INDENT 0.0
.INDENT 3.5
.sp
.nf
.ft C
Label: \(aqBTRFS\(aq  uuid: 8a9d72cd\-ead3\-469d\-b371\-9c7203276265
        Total devices 2 FS bytes used 77.03GiB
        devid    1 size 53.90GiB used 51.90GiB path /dev/sdc2
        devid    2 size 53.90GiB used 51.90GiB path /dev/sde1
.ft P
.fi
.UNINDENT
.UNINDENT
.sp
\fIsize\fP \- \fIused\fP = \fIfree work space\fP
.sp
\fI53.90GiB\fP \- \fI51.90GiB\fP = \fI2.00GiB\fP
.sp
An example of a filter that does not require workspace is \fIusage=0\fP\&. This will
scan through all unused block groups of a given type and will reclaim the
space. After that it might be possible to run other filters.
.sp
\fBCONVERSIONS ON MULTIPLE DEVICES\fP
.sp
Conversion to profiles based on striping (RAID0, RAID5/6) require the work
space on each device. An interrupted balance may leave partially filled block
groups that consume the work space.
.SH EXAMPLES
.sp
A more comprehensive example when going from one to multiple devices, and back,
can be found in section \fITYPICAL USECASES\fP of \fI\%btrfs\-device(8)\fP\&.
.SS MAKING BLOCK GROUP LAYOUT MORE COMPACT
.sp
The layout of block groups is not normally visible; most tools report only
summarized numbers of free or used space, but there are still some hints
provided.
.sp
Let\(aqs use the following real life example and start with the output:
.INDENT 0.0
.INDENT 3.5
.sp
.nf
.ft C
$ btrfs filesystem df /path
Data, single: total=75.81GiB, used=64.44GiB
System, RAID1: total=32.00MiB, used=20.00KiB
Metadata, RAID1: total=15.87GiB, used=8.84GiB
GlobalReserve, single: total=512.00MiB, used=0.00B
.ft P
.fi
.UNINDENT
.UNINDENT
.sp
Roughly calculating for data, \fI75G \- 64G = 11G\fP, the used/total ratio is
about \fI85%\fP\&. How can we can interpret that:
.INDENT 0.0
.IP \(bu 2
chunks are filled by 85% on average, i.e. the \fIusage\fP filter with anything
smaller than 85 will likely not affect anything
.IP \(bu 2
in a more realistic scenario, the space is distributed unevenly, we can
assume there are completely used chunks and the remaining are partially filled
.UNINDENT
.sp
Compacting the layout could be used on both. In the former case it would spread
data of a given chunk to the others and removing it. Here we can estimate that
roughly 850 MiB of data have to be moved (85% of a 1 GiB chunk).
.sp
In the latter case, targeting the partially used chunks will have to move less
data and thus will be faster. A typical filter command would look like:
.INDENT 0.0
.INDENT 3.5
.sp
.nf
.ft C
# btrfs balance start \-dusage=50 /path
Done, had to relocate 2 out of 97 chunks

$ btrfs filesystem df /path
Data, single: total=74.03GiB, used=64.43GiB
System, RAID1: total=32.00MiB, used=20.00KiB
Metadata, RAID1: total=15.87GiB, used=8.84GiB
GlobalReserve, single: total=512.00MiB, used=0.00B
.ft P
.fi
.UNINDENT
.UNINDENT
.sp
As you can see, the \fItotal\fP amount of data is decreased by just 1 GiB, which is
an expected result. Let\(aqs see what will happen when we increase the estimated
usage filter.
.INDENT 0.0
.INDENT 3.5
.sp
.nf
.ft C
# btrfs balance start \-dusage=85 /path
Done, had to relocate 13 out of 95 chunks

$ btrfs filesystem df /path
Data, single: total=68.03GiB, used=64.43GiB
System, RAID1: total=32.00MiB, used=20.00KiB
Metadata, RAID1: total=15.87GiB, used=8.85GiB
GlobalReserve, single: total=512.00MiB, used=0.00B
.ft P
.fi
.UNINDENT
.UNINDENT
.sp
Now the used/total ratio is about 94% and we moved about \fI74G \- 68G = 6G\fP of
data to the remaining block groups, i.e. the 6GiB are now free of filesystem
structures, and can be reused for new data or metadata block groups.
.sp
We can do a similar exercise with the metadata block groups, but this should
not typically be necessary, unless the used/total ratio is really off. Here
the ratio is roughly 50% but the difference as an absolute number is \(dqa few
gigabytes\(dq, which can be considered normal for a workload with snapshots or
reflinks updated frequently.
.INDENT 0.0
.INDENT 3.5
.sp
.nf
.ft C
# btrfs balance start \-musage=50 /path
Done, had to relocate 4 out of 89 chunks

$ btrfs filesystem df /path
Data, single: total=68.03GiB, used=64.43GiB
System, RAID1: total=32.00MiB, used=20.00KiB
Metadata, RAID1: total=14.87GiB, used=8.85GiB
GlobalReserve, single: total=512.00MiB, used=0.00B
.ft P
.fi
.UNINDENT
.UNINDENT
.sp
Just 1 GiB decrease, which possibly means there are block groups with good
utilization. Making the metadata layout more compact would in turn require
updating more metadata structures, i.e. lots of IO. As running out of metadata
space is a more severe problem, it\(aqs not necessary to keep the utilization
ratio too high. For the purpose of this example, let\(aqs see the effects of
further compaction:
.INDENT 0.0
.INDENT 3.5
.sp
.nf
.ft C
# btrfs balance start \-musage=70 /path
Done, had to relocate 13 out of 88 chunks

$ btrfs filesystem df .
Data, single: total=68.03GiB, used=64.43GiB
System, RAID1: total=32.00MiB, used=20.00KiB
Metadata, RAID1: total=11.97GiB, used=8.83GiB
GlobalReserve, single: total=512.00MiB, used=0.00B
.ft P
.fi
.UNINDENT
.UNINDENT
.SS GETTING RID OF COMPLETELY UNUSED BLOCK GROUPS
.sp
Normally the balance operation needs a work space, to temporarily move the
data before the old block groups gets removed. If there\(aqs no work space, it
ends with \fIno space left\fP\&.
.sp
There\(aqs a special case when the block groups are completely unused, possibly
left after removing lots of files or deleting snapshots. Removing empty block
groups is automatic since 3.18. The same can be achieved manually with a
notable exception that this operation does not require the work space. Thus it
can be used to reclaim unused block groups to make it available.
.INDENT 0.0
.INDENT 3.5
.sp
.nf
.ft C
# btrfs balance start \-dusage=0 /path
.ft P
.fi
.UNINDENT
.UNINDENT
.sp
This should lead to decrease in the \fItotal\fP numbers in the \fBbtrfs filesystem df\fP output.
.SH EXIT STATUS
.sp
Unless indicated otherwise below, all \fBbtrfs balance\fP subcommands
return a zero exit status if they succeed, and non zero in case of
failure.
.sp
The \fBpause\fP, \fBcancel\fP, and \fBresume\fP subcommands exit with a status of
\fB2\fP if they fail because a balance operation was not running.
.sp
The \fBstatus\fP subcommand exits with a status of \fB0\fP if a balance
operation is not running, \fB1\fP if the command\-line usage is incorrect
or a balance operation is still running, and \fB2\fP on other errors.
.SH AVAILABILITY
.sp
\fBbtrfs\fP is part of btrfs\-progs.  Please refer to the documentation at
\fI\%https://btrfs.readthedocs.io\fP or wiki \fI\%http://btrfs.wiki.kernel.org\fP for further
information.
.SH SEE ALSO
.sp
\fI\%mkfs.btrfs(8)\fP,
\fI\%btrfs\-device(8)\fP
.\" Generated by docutils manpage writer.
.