LVM.CONF(5) File Formats Manual LVM.CONF(5)
NAME
lvm.conf - Configuration file for LVM2
SYNOPSIS
/etc/lvm/lvm.conf
DESCRIPTION
lvm.conf is loaded during the initialisation phase of lvm (8). This file can in turn lead to other files being loaded - settings read in
later override earlier settings. File timestamps are checked between commands and if any have changed, all the files are reloaded.
Use lvm dumpconfig to check what settings are in use.
SYNTAX
This section describes the configuration file syntax.
Whitespace is not significant unless it is within quotes. This provides a wide choice of acceptable indentation styles. Comments begin
with # and continue to the end of the line. They are treated as whitespace.
Here is an informal grammar:
file = value*
A configuration file consists of a set of values.
value = section | assignment
A value can either be a new section, or an assignment.
section = identifier '{' value* '}'
A section is groups associated values together.
It is denoted by a name and delimited by curly brackets.
e.g. backup {
...
}
assignment = identifier '=' (array | type)
An assignment associates a type with an identifier.
e.g. max_archives = 42
array = '[' (type ',')* type ']' | '[' ']'
Inhomogeneous arrays are supported.
Elements must be separated by commas.
An empty array is acceptable.
type = integer | float | string
integer = [0-9]*
float = [0-9]*.[0-9]*
string = '"' .* '"'
Strings must be enclosed in double quotes.
SECTIONS
The sections that may be present in the file are:
devices -- Device settings
dir -- Directory in which to create volume group device nodes. Defaults to "/dev". Commands also accept this as a prefix on volume
group names.
scan -- List of directories to scan recursively for LVM physical volumes. Devices in directories outside this hierarchy will be
ignored. Defaults to "/dev".
preferred_names -- List of patterns compared in turn against all the pathnames referencing the same device in in the scanned direc-
tories. The pathname that matches the earliest pattern in the list is the one used in any output. As an example, if device-mapper
multipathing is used, the following will select multipath device names:
devices { preferred_names = [ "^/dev/mapper/mpath" ] }
filter -- List of patterns to apply to devices found by a scan. Patterns are regular expressions delimited by any character and
preceded by a (for accept) or r (for reject). The list is traversed in order, and the first regex that matches determines if the
device will be accepted or rejected (ignored). Devices that don't match any patterns are accepted. If you want to reject patterns
that don't match, end the list with "r/.*/". If there are several names for the same device (e.g. symbolic links in /dev), if the
first matching pattern in the list for any of the names is an a pattern, the device is accepted; otherwise if the first matching
pattern in the list for any of the names is an r pattern it is rejected; otherwise it is accepted. As an example, to ignore
/dev/cdrom you could use:
devices { filter=["r|cdrom|"] }
cache_dir -- Persistent filter cache file directory. Defaults to "/etc/lvm/cache".
write_cache_state -- Set to 0 to disable the writing out of the persistent filter cache file when lvm exits. Defaults to 1.
types -- List of pairs of additional acceptable block device types found in /proc/devices together with maximum (non-zero) number of
partitions (normally 16). By default, LVM2 supports ide, sd, md, loop, dasd, dac960, nbd, ida, cciss, ubd, ataraid, drbd, power2,
i2o_block and iseries/vd. Block devices with major numbers of different types are ignored by LVM2. Example: types = ["fd", 16].
To create physical volumes on device-mapper volumes created outside LVM2, perhaps encrypted ones from cryptsetup, you'll need types
= ["device-mapper", 16]. But if you do this, be careful to avoid recursion within LVM2. The figure for number of partitions is not
currently used in LVM2 - and might never be.
sysfs_scan -- If set to 1 and your kernel supports sysfs and it is mounted, sysfs will be used as a quick way of filtering out block
devices that are not present.
md_component_detection -- If set to 1, LVM2 will ignore devices used as components of software RAID (md) devices by looking for md
superblocks. This doesn't always work satisfactorily e.g. if a device has been reused without wiping the md superblocks first.
md_chunk_alignment -- If set to 1, and a Physical Volume is placed directly upon an md device, LVM2 will align its data blocks with
the md device's stripe-width.
data_alignment_detection -- If set to 1, and your kernel provides topology information in sysfs for the Physical Volume, the start
of data area will be aligned on a multiple of the 'minimum_io_size' or 'optimal_io_size' exposed in sysfs. minimum_io_size is the
smallest request the device can perform without incurring a read-modify-write penalty (e.g. MD's chunk size). optimal_io_size is
the device's preferred unit of receiving I/O (e.g. MD's stripe width). minimum_io_size is used if optimal_io_size is undefined (0).
If both md_chunk_alignment and data_alignment_detection are enabled the result of data_alignment_detection is used.
data_alignment -- Default alignment (in KB) of start of data area when creating a new Physical Volume using the lvm2 format. If a
Physical Volume is placed directly upon an md device and md_chunk_alignment or data_alignment_detection is enabled this parameter is
ignored. Set to 0 to use the default alignment of 64KB or the page size, if larger.
data_alignment_offset_detection -- If set to 1, and your kernel provides topology information in sysfs for the Physical Volume, the
start of the aligned data area of the Physical Volume will be shifted by the alignment_offset exposed in sysfs.
To see the location of the first Physical Extent of an existing Physical Volume use pvs -o +pe_start . It will be a multiple of the
requested data_alignment plus the alignment_offset from data_alignment_offset_detection (if enabled) or the pvcreate commandline.
disable_after_error_count -- During each LVM operation errors received from each device are counted. If the counter of a particular
device exceeds the limit set here, no further I/O is sent to that device for the remainder of the respective operation. Setting the
parameter to 0 disables the counters altogether.
pv_min_size -- Minimal size (in KB) of the block device which can be used as a PV. In clustered environment all nodes have to use
the same value. Any value smaller than 512KB is ignored. Up to and include version 2.02.84 the default was 512KB. From 2.02.85
onwards it was changed to 2MB to avoid floppy drives by default.
issue_discards -- Issue discards to a logical volumes's underlying physical volume(s) when the logical volume is no longer using the
physical volumes' space (e.g. lvremove, lvreduce, etc). Discards inform the storage that a region is no longer in use. Storage
that supports discards advertise the protocol specific way discards should be issued by the kernel (TRIM, UNMAP, or WRITE SAME with
UNMAP bit set). Not all storage will support or benefit from discards but SSDs and thinly provisioned LUNs generally do. If set to
1, discards will only be issued if both the storage and kernel provide support.
allocation -- Space allocation policies
cling_tag_list -- List of PV tags matched by the cling allocation policy.
When searching for free space to extend an LV, the cling allocation policy will choose space on the same PVs as the last segment of
the existing LV. If there is insufficient space and a list of tags is defined here, it will check whether any of them are attached
to the PVs concerned and then seek to match those PV tags between existing extents and new extents.
The @ prefix for tags is required. Use the special tag "@*" as a wildcard to match any PV tag and so use all PV tags for this pur-
pose.
For example, LVs are mirrored between two sites within a single VG. PVs are tagged with either @site1 or @site2 to indicate where
they are situated and these two PV tags are selected for use with this allocation policy:
cling_tag_list = [ "@site1", "@site2" ]
log -- Default log settings
file -- Location of log file. If this entry is not present, no log file is written.
overwrite -- Set to 1 to overwrite the log file each time a tool is invoked. By default tools append messages to the log file.
level -- Log level (0-9) of messages to write to the file. 9 is the most verbose; 0 should produce no output.
verbose -- Default level (0-3) of messages sent to stdout or stderr. 3 is the most verbose; 0 should produce the least output.
syslog -- Set to 1 (the default) to send log messages through syslog. Turn off by setting to 0. If you set to an integer greater
than one, this is used - unvalidated - as the facility. The default is LOG_USER. See /usr/include/sys/syslog.h for safe facility
values to use. For example, LOG_LOCAL0 might be 128.
indent -- When set to 1 (the default) messages are indented according to their severity, two spaces per level. Set to 0 to turn off
indentation.
command_names -- When set to 1, the command name is used as a prefix for each message. Default is 0 (off).
prefix -- Prefix used for all messages (after the command name). Default is two spaces.
activation -- Set to 1 to log messages while devices are suspended during activation. Only set this temporarily while debugging a
problem because in low memory situations this setting can cause your machine to lock up.
backup -- Configuration for metadata backups.
archive_dir -- Directory used for automatic metadata archives. Backup copies of former metadata for each volume group are archived
here. Defaults to "/etc/lvm/archive".
backup_dir -- Directory used for automatic metadata backups. A single backup copy of the current metadata for each volume group is
stored here. Defaults to "/etc/lvm/backup".
archive -- Whether or not tools automatically archive existing metadata into archive_dir before making changes to it. Default is 1
(automatic archives enabled). Set to 0 to disable. Disabling this might make metadata recovery difficult or impossible if some-
thing goes wrong.
backup -- Whether or not tools make an automatic backup into backup_dir after changing metadata. Default is 1 (automatic backups
enabled). Set to 0 to disable. Disabling this might make metadata recovery difficult or impossible if something goes wrong.
retain_min -- Minimum number of archives to keep. Defaults to 10.
retain_days -- Minimum number of days to keep archive files. Defaults to 30.
shell -- LVM2 built-in readline shell settings
history_size -- Maximum number of lines of shell history to retain (default 100) in $HOME/.lvm_history
global -- Global settings
test -- If set to 1, run tools in test mode i.e. no changes to the on-disk metadata will get made. It's equivalent to having the -t
option on every command.
activation -- Set to 0 to turn off all communication with the device-mapper driver. Useful if you want to manipulate logical vol-
umes while device-mapper is not present in your kernel.
proc -- Mount point of proc filesystem. Defaults to /proc.
umask -- File creation mask for any files and directories created. Interpreted as octal if the first digit is zero. Defaults to
077. Use 022 to allow other users to read the files by default.
format -- The default value of --metadatatype used to determine which format of metadata to use when creating new physical volumes
and volume groups. lvm1 or lvm2.
fallback_to_lvm1 -- Set this to 1 if you need to be able to switch between 2.4 kernels using LVM1 and kernels including device-map-
per. The LVM2 tools should be installed as normal and the LVM1 tools should be installed with a .lvm1 suffix e.g. vgscan.lvm1. If
an LVM2 tool is then run but unable to communicate with device-mapper, it will automatically invoke the equivalent LVM1 version of
the tool. Note that for LVM1 tools to manipulate physical volumes and volume groups created by LVM2 you must use --metadataformat
lvm1 when creating them.
library_dir -- A directory searched for LVM2's shared libraries ahead of the places dlopen (3) searches.
format_libraries -- A list of shared libraries to load that contain code to process different formats of metadata. For example,
liblvm2formatpool.so is needed to read GFS pool metadata if LVM2 was configured --with-pool=shared.
locking_type -- What type of locking to use. 1 is the default, which use flocks on files in locking_dir (see below) to avoid con-
flicting LVM2 commands running concurrently on a single machine. 0 disables locking and risks corrupting your metadata. If set to
2, the tools will load the external locking_library (see below). If the tools were configured --with-cluster=internal (the default)
then 3 means to use built-in cluster-wide locking. Type 4 enforces read-only metadata and forbids any operations that might want to
modify Volume Group metadata. All changes to logical volumes and their states are communicated using locks.
wait_for_locks -- When set to 1, the default, the tools wait if a lock request cannot be satisfied immediately. When set to 0, the
operation is aborted instead.
locking_dir -- The directory LVM2 places its file locks if locking_type is set to 1. The default is /var/lock/lvm.
locking_library -- The name of the external locking library to load if locking_type is set to 2. The default is liblvm2cluster-
lock.so. If you need to write such a library, look at the lib/locking source code directory.
tags -- Host tag settings
hosttags -- If set to 1, create a host tag with the machine name. Setting this to 0 does nothing, neither creating nor destroying
any tag. The machine name used is the nodename as returned by uname (2).
Additional host tags to be set can be listed here as subsections. The @ prefix for tags is optional. Each of these host tag sub-
sections can contain a host_list array of host names. If any one of these entries matches the machine name exactly then the host tag
gets defined on this particular host, otherwise it doesn't.
After lvm.conf has been processed, LVM2 works through each host tag that has been defined in turn, and if there is a configuration
file called lvm_<host_tag>.conf it attempts to load it. Any settings read in override settings found in earlier files. Any addi-
tional host tags defined get appended to the search list, so in turn they can lead to further configuration files being processed.
Use lvm dumpconfig to check the result of config file processing.
The following example always sets host tags tag1 and sets tag2 on machines fs1 and fs2:
tags { tag1 { } tag2 { host_list = [ "fs1", "fs2" ] } }
These options are useful if you are replicating configuration files around a cluster. Use of hosttags = 1 means every machine can
have static and identical local configuration files yet use different settings and activate different logical volumes by default.
See also volume_list below and --addtag in lvm (8).
activation -- Settings affecting device-mapper activation
missing_stripe_filler -- When activating an incomplete logical volume in partial mode, this option dictates how the missing data is
replaced. A value of "error" will cause activation to create error mappings for the missing data, meaning that read access to miss-
ing portions of the volume will result in I/O errors. You can instead also use a device path, and in that case this device will be
used in place of missing stripes. However, note that using anything other than "error" with mirrored or snapshotted volumes is
likely to result in data corruption. For instructions on how to create a device that always returns zeros, see lvcreate (8).
mirror_region_size -- Unit size in KB for copy operations when mirroring.
readahead -- Used when there is no readahead value stored in the volume group metadata. Set to none to disable readahead in these
circumstances or auto to use the default value chosen by the kernel.
reserved_memory, reserved_stack -- How many KB to reserve for LVM2 to use while logical volumes are suspended. If insufficient mem-
ory is reserved before suspension, there is a risk of machine deadlock.
process_priority -- The nice value to use while devices are suspended. This is set to a high priority so that logical volumes are
suspended (with I/O generated by other processes to those logical volumes getting queued) for the shortest possible time.
volume_list -- This acts as a filter through which all requests to activate a logical volume on this machine are passed. A logical
volume is only activated if it matches an item in the list. Tags must be preceded by @ and are checked against all tags defined in
the logical volume and volume group metadata for a match. @* is short-hand to check every tag set on the host machine (see tags
above). Logical volume and volume groups can also be included in the list by name e.g. vg00, vg00/lvol1.
read_only_volume_list -- This acts as a filter through which all requests to activate a logical volume on this machine are passed.
A logical volume is activated in read-only mode (instead of read-write) if it matches an item in the list. Volumes must first pass
the volume_list filter, if present. Tags must be preceded by @ and are checked against all tags defined in the logical volume and
volume group metadata for a match. @* is short-hand to check every tag set on the host machine (see tags above). Logical volume
and volume groups can also be included in the list by name e.g. vg00, vg00/lvol1.
metadata -- Advanced metadata settings
pvmetadatacopies -- When creating a physical volume using the LVM2 metadata format, this is the default number of copies of metadata
to store on each physical volume. Currently it can be set to 0, 1 or 2. The default is 1. If set to 2, one copy is placed at the
beginning of the disk and the other is placed at the end. It can be overridden on the command line with --pvmetadatacopies (see
pvcreate). If creating a volume group with just one physical volume, it's a good idea to have 2 copies. If creating a large volume
group with many physical volumes, you may decide that 3 copies of the metadata is sufficient, i.e. setting it to 1 on three of the
physical volumes, and 0 on the rest. Every volume group must contain at least one physical volume with at least 1 copy of the meta-
data (unless using the text files described below). The disadvantage of having lots of copies is that every time the tools access
the volume group, every copy of the metadata has to be accessed, and this slows down the tools.
pvmetadatasize -- Approximate number of sectors to set aside for each copy of the metadata. Volume groups with large numbers of
physical or logical volumes, or volumes groups containing complex logical volume structures will need additional space for their
metadata. The metadata areas are treated as circular buffers, so unused space becomes filled with an archive of the most recent
previous versions of the metadata.
pvmetadataignore When creating a physical volume using the LVM2 metadata format, this states whether metadata areas should be
ignored. The default is "n". If metadata areas on a physical volume are ignored, LVM will not not store metadata in the metadata
areas present on newly created Physical Volumes. The option can be overridden on the command line with --metadataignore (See pvcre-
ate and pvchange). Metadata areas cannot be created or extended after Logical Volumes have been allocated on the device. If you do
not want to store metadata on this device, it is still wise always to allocate a metadata area (use a non-zero value for --pvmeta-
datacopies) in case you need it in the future and to use this option to instruct LVM2 to ignore it.
vgmetadatacopies -- When creating a volume group using the LVM2 metadata format, this is the default number of copies of metadata
desired across all the physical volumes in the volume group. If set to a non-zero value, LVM will automatically set or clear the
metadataignore flag on the physical volumes (see pvcreate and pvchange --metadataignore) in order to achieve the desired number of
metadata copies. An LVM command that adds or removes physical volumes (for example, vgextend, vgreduce, vgsplit, or vgmerge), may
cause LVM to automatically set or clear the metadataignore flags. Also, if physical volumes go missing or reappear, or a new number
of copies is explicitly set (see vgchange --vgmetadatacopies), LVM may adjust the metadataignore flags. Set vgmetadatacopies to 0
instructs LVM not to set or clear the metadataignore flags automatically. You may set a value larger than the sum of all metadata
areas on all physical volumes. The value can be overridden on the command line with --vgmetadatacopies for various commands (for
example, vgcreate and vgchange), and can be queryied with the vg_mda_copies field of vgs. This option is useful for volume groups
containing large numbers of physical volumes with metadata as it may be used to minimize metadata read and write overhead.
dirs -- List of directories holding live copies of LVM2 metadata as text files. These directories must not be on logical volumes.
It is possible to use LVM2 with a couple of directories here, preferably on different (non-logical-volume) filesystems and with no
other on-disk metadata, pvmetadatacopies = 0. Alternatively these directories can be in addition to the on-disk metadata areas.
This feature was created during the development of the LVM2 metadata before the new on-disk metadata areas were designed and no
longer gets tested. It is not supported under low-memory conditions, and it is important never to edit these metadata files unless
you fully understand how things work: to make changes you should always use the tools as normal, or else vgcfgbackup, edit backup,
vgcfgrestore.
FILES
/etc/lvm/lvm.conf /etc/lvm/archive /etc/lvm/backup /etc/lvm/cache/.cache /run/lock/lvm
SEE ALSO
lvm(8), umask(2), uname(2), dlopen(3), syslog(3), syslog.conf(5)
Sistina Software UK LVM TOOLS 2.02.95(2) (2012-03-06) LVM.CONF(5)