2. EL 8
  3. system-storage-manager
  4. ssm(8)

Scroll to navigation

SSM(8) System Storage Manager SSM(8)

NAME

ssm - System Storage Manager: a single tool to manage your storage

SYNOPSIS

ssm [-h] [--version] [-v] [-v**v] [-v**vv] [-f] [-b BACKEND] [-n] {check,resize,create,list,add,remove,snapshot,mount,migrate} ...

ssm create [-h] [-s SIZE] [-n NAME] [--fstype FSTYPE] [-r LEVEL] [-I STRIPESIZE] [-i STRIPES] [-p POOL] [-e [{luks,plain}]] [-o MNT_OPTIONS] [-v VIRTUAL_SIZE] [device [device ...]] [mount]

ssm list [-h] [{volumes,vol,dev,devices,pool,pools,fs,filesystems,snap,snapshots}]

ssm remove [-h] [-a] [items [items ...]]

ssm resize [-h] [-s SIZE] volume [device [device ...]]

ssm check [-h] device [device ...]

ssm snapshot [-h] [-s SIZE] [-d DEST | -n NAME] volume

ssm add [-h] [-p POOL] device [device ...]

ssm mount [-h] [-o OPTIONS] volume directory

ssm migrate [-h] source target

DESCRIPTION

System Storage Manager provides an easy to use command line interface to manage your storage using various technologies like lvm, btrfs, encrypted volumes and more.

In more sophisticated enterprise storage environments, management with Device Mapper (dm), Logical Volume Manager (LVM), or Multiple Devices (md) is becoming increasingly more difficult. With file systems added to the mix, the number of tools needed to configure and manage storage has grown so large that it is simply not user friendly. With so many options for a system administrator to consider, the opportunity for errors and problems is large.

The btrfs administration tools have shown us that storage management can be simplified, and we are working to bring that ease of use to Linux filesystems in general.

OPTIONS

show this help message and exit
show program's version number and exit
-v, --verbose
Show aditional information while executing.
-vv
Show yet more aditional information while executing.
-vvv
Show yet more aditional information while executing.
Force execution in the case where ssm has some doubts or questions.
Choose backend to use. Currently you can choose from (lvm,btrfs,crypt).
-n, --dry-run
Dry run. Do not do anything, just parse the command line options and gather system information if necessary. Note that with this option ssm will not perform all the check as some of them are done by the backends themselves. This option is mainly used for debugging purposes, but still requires root privileges.

SYSTEM STORAGE MANAGER COMMANDS

Introduction

System Storage Manager has several commands that you can specify on the command line as a first argument to ssm. They all have a specific use and their own arguments, but global ssm arguments are propagated to all commands.

Create command

ssm create [-h] [-s SIZE] [-n NAME] [--fstype FSTYPE] [-r LEVEL] [-I STRIPESIZE] [-i STRIPES] [-p POOL] [-e [{luks,plain}]] [-o MNT_OPTIONS] [-v VIRTUAL_SIZE] [device [device ...]] [mount]

This command creates a new volume with defined parameters. If a device is provided it will be used to create the volume, hence it will be added into the pool prior to volume creation (See Add command section). More than one device can be used to create a volume.

If the device is already being used in a different pool, then ssm will ask you whether you want to remove it from the original pool. If you decline, or the removal fails, then the volume creation fails if the SIZE was not provided. On the other hand, if the SIZE is provided and some devices can not be added to the pool, the volume creation might still succeed if there is enough space in the pool.

In addition to specifying size of the volume directly, percentage can be specified as well. Specify --size 70% to indicate the volume size to be 70% of total pool size. Additionally, percentage of the used, or free pool space can be specified as well using keywords FREE, or USED respectively.

The POOL name can be specified as well. If the pool exists, a new volume will be created from that pool (optionally adding device into the pool). However if the POOL does not exist, then ssm will attempt to create a new pool with the provided device, and then create a new volume from this pool. If the --backend argument is omitted, the default ssm backend will be used. The default backend is lvm.

ssm also supports creating a RAID configuration, however some back-ends might not support all RAID levels, or may not even support RAID at all. In this case, volume creation will fail.

If a mount point is provided, ssm will attempt to mount the volume after it is created. However it will fail if mountable file system is not present on the volume.

If the backend allows it (currently only supported with lvm backend), ssm can be used to create thinly provisioned volumes by specifying --virtual-size option. This will automatically create a thin pool of a given size provided with --size option and thin volume of a given size provided with --virtual-size option and name provided with --name option. Virtual size can be much bigger than available space in the pool.

show this help message and exit
Gives the size to allocate for the new logical volume. A size suffix K|k, M|m, G|g, T|t, P|p, E|e can be used to define 'power of two' units. If no unit is provided, it defaults to kilobytes. This is optional and if not given, maximum possible size will be used. Additionally the new size can be specified as a percentage of the total pool size (50%), as a percentage of free pool space (50%FREE), or as a percentage of used pool space (50%USED).
The name for the new logical volume. This is optional and if omitted, name will be generated by the corresponding backend.
Gives the file system type to create on the new logical volume. Supported file systems are (ext3, ext4, xfs, btrfs). This is optional and if not given file system will not be created.
Specify a RAID level you want to use when creating a new volume. Note that some backends might not implement all supported RAID levels. This is optional and if no specified, linear volume will be created. You can choose from the following list of supported levels (0,1,10).
Gives the number of kilobytes for the granularity of stripes. This is optional and if not given, backend default will be used. Note that you have to specify RAID level as well.
Gives the number of stripes. This is equal to the number of physical volumes to scatter the logical volume. This is optional and if stripesize is set and multiple devices are provided stripes is determined automatically from the number of devices. Note that you have to specify RAID level as well.
Pool to use to create the new volume.

Create encrpted volume. Extension to use can be specified.

Mount options are specified with a -o flag followed by a comma separated string of options. This option is equivalent to the -o mount(8) option.
Gives the virtual size for the new thinly provisioned volume. A size suffix K|k, M|m, G|g, T|t, P|p, E|e can be used to define 'power of two' units. If no unit is provided, it defaults to kilobytes.

List command

ssm list [-h] [{volumes,vol,dev,devices,pool,pools,fs,filesystems,snap,snapshots}]

Lists information about all detected devices, pools, volumes and snapshots found on the system. The list command can be used either alone to list all of the information, or you can request specific sections only.

The following sections can be specified:

{volumes | vol}
List information about all volumes found in the system.
{devices | dev}
List information about all devices found on the system. Some devices are intentionally hidden, like for example cdrom or DM/MD devices since those are actually listed as volumes.
{pools | pool}
List information about all pools found in the system.
{filesystems | fs}
List information about all volumes containing filesystems found in the system.
{snapshots | snap}
List information about all snapshots found in the system. Note that some back-ends do not support snapshotting and some cannot distinguish snapshot from regular volumes. In this case, ssm will try to recognize the volume name in order to identify a snapshot, but if the ssm regular expression does not match the snapshot pattern, the problematic snapshot will not be recognized.

show this help message and exit

Remove command

ssm remove [-h] [-a] [items [items ...]]

This command removes an item from the system. Multiple items can be specified. If the item cannot be removed for some reason, it will be skipped.

An item can be any of the following:

Remove a device from the pool. Note that this cannot be done in some cases where the device is being used by the pool. You can use the -f argument to force removal. If the device does not belong to any pool, it will be skipped.
Remove a pool from the system. This will also remove all volumes created from that pool.
Remove a volume from the system. Note that this will fail if the volume is mounted and cannot be forced with -f.

show this help message and exit
Remove all pools in the system.

Resize command

ssm resize [-h] [-s SIZE] volume [device [device ...]]

Change size of the volume and file system. If there is no file system, only the volume itself will be resized. You can specify a device to add into the volume pool prior the resize. Note that the device will only be added into the pool if the volume size is going to grow.

If the device is already used in a different pool, then ssm will ask you whether or not you want to remove it from the original pool.

In some cases, the file system has to be mounted in order to resize. This will be handled by ssm automatically by mounting the volume temporarily.

In addition to specifying new size of the volume directly, percentage can be specified as well. Specify --size 70% to resize the volume to 70% of it's original size. Additionally, percentage of the used, or free pool space can be specified as well using keywords FREE, or USED respectively.

Note that resizing btrfs subvolume is not supported, only the whole file system can be resized.

show this help message and exit
New size of the volume. With the + or - sign the value is added to or subtracted from the actual size of the volume and without it, the value will be set as the new volume size. A size suffix of [k|K] for kilobytes, [m|M] for megabytes, [g|G] for gigabytes, [t|T] for terabytes or [p|P] for petabytes is optional. If no unit is provided the default is kilobytes. Additionally the new size can be specified as a percentage of the original volume size ([+][-]50%), as a percentage of free pool space ([+][-]50%FREE), or as a percentage of used pool space ([+][-]50%USED).

Check command

ssm check [-h] device [device ...]

Check the file system consistency on the volume. You can specify multiple volumes to check. If there is no file system on the volume, this volume will be skipped.

In some cases the file system has to be mounted in order to check the file system. This will be handled by ssm automatically by mounting the volume temporarily.

show this help message and exit

Snapshot command

ssm snapshot [-h] [-s SIZE] [-d DEST | -n NAME] volume

Take a snapshot of an existing volume. This operation will fail if the back-end to which the volume belongs to does not support snapshotting. Note that you cannot specify both NAME and DEST since those options are mutually exclusive.

In addition to specifying new size of the volume directly, percentage can be specified as well. Specify --size 70% to indicate the new snapshot size to be 70% of the origin volume size. Additionally, percentage of the used, or free pool space can be specified as well using keywords FREE, or USED respectively.

In some cases the file system has to be mounted in order to take a snapshot of the volume. This will be handled by ssm automatically by mounting the volume temporarily.

show this help message and exit
Gives the size to allocate for the new snapshot volume. A size suffix K|k, M|m, G|g, T|t, P|p, E|e can be used to define 'power of two' units. If no unit is provided, it defaults to kilobytes. This is optional and if not given, the size will be determined automatically. Additionally the new size can be specified as a percentage of the original volume size (50%), as a percentage of free pool space (50%FREE), or as a percentage of used pool space (50%USED).
Destination of the snapshot specified with absolute path to be used for the new snapshot. This is optional and if not specified default backend policy will be performed.
Name of the new snapshot. This is optional and if not specified default backend policy will be performed.

Add command

ssm add [-h] [-p POOL] device [device ...]

This command adds a device into the pool. By default, the device will not be added if it's already a part of a different pool, but the user will be asked whether or not to remove the device from its pool. When multiple devices are provided, all of them are added into the pool. If one of the devices cannot be added into the pool for any reason, the add command will fail. If no pool is specified, the default pool will be chosen. In the case of a non existing pool, it will be created using the provided devices.

show this help message and exit
Pool to add device into. If not specified the default pool is used.

Mount command

ssm mount [-h] [-o OPTIONS] volume directory

This command will mount the volume at the specified directory. The volume can be specified in the same way as with mount(8), however in addition, one can also specify a volume in the format as it appears in the ssm list table.

For example, instead of finding out what the device and subvolume id of the btrfs subvolume "btrfs_pool:vol001" is in order to mount it, one can simply call ssm mount btrfs_pool:vol001 /mnt/test.

One can also specify OPTIONS in the same way as with mount(8).

show this help message and exit
Options are specified with a -o flag followed by a comma separated string of options. This option is equivalent to the same mount(8) option.

Migrate command

ssm migrate [-h] source target

Move data from one device to another. For btrfs and lvm their specialized utilities are used, so the data are moved in an all-or-nothing fashion and no other operation is needed to add/remove the devices or rebalance the pool. Devices that do not belong to a backend that supports specialized device migration tools will be migrated using dd.

This operation is not intended to be used for duplication, because the process can change metadata and an access to the data may be difficult.

show this help message and exit

BACK-ENDS

Introduction

Ssm aims to create a unified user interface for various technologies like Device Mapper (dm), Btrfs file system, Multiple Devices (md) and possibly more. In order to do so we have a core abstraction layer in ssmlib/main.py. This abstraction layer should ideally know nothing about the underlying technology, but rather comply with device, pool and volume abstractions.

Various backends can be registered in ssmlib/main.py in order to handle specific storage technology, implementing methods like create, snapshot, or remove volumes and pools. The core will then call these methods to manage the storage without needing to know what lies underneath it. There are already several backends registered in ssm.

Btrfs backend

Btrfs is the file system with many advanced features including volume management. This is the reason why btrfs is handled differently than other conventional file systems in ssm. It is used as a volume management back-end.

Pools, volumes and snapshots can be created with btrfs backend and here is what it means from the btrfs point of view:

A pool is actually a btrfs file system itself, because it can be extended by adding more devices, or shrunk by removing devices from it. Subvolumes and snapshots can also be created. When the new btrfs pool should be created, ssm simply creates a btrfs file system, which means that every new btrfs pool has one volume of the same name as the pool itself which can not be removed without removing the entire pool. The default btrfs pool name is btrfs_pool.

When creating a new btrfs pool, the name of the pool is used as the file system label. If there is an already existing btrfs file system in the system without a label, a btrfs pool name will be generated for internal use in the following format "btrfs_{device base name}".

A btrfs pool is created when the create or add command is used with specified devices and non existing pool name.

A volume in the btrfs back-end is actually just btrfs subvolume with the exception of the first volume created on btrfs pool creation, which is the file system itself. Subvolumes can only be created on the btrfs file system when it is mounted, but the user does not have to worry about that since ssm will automatically mount the file system temporarily in order to create a new subvolume.

The volume name is used as subvolume path in the btrfs file system and every object in this path must exist in order to create a volume. The volume name for internal tracking and that is visible to the user is generated in the format "{pool_name}:{volume name}", but volumes can be also referenced by its mount point.

The btrfs volumes are only shown in the list output, when the file system is mounted, with the exception of the main btrfs volume - the file system itself.

Also note that btrfs volumes and subvolumes cannot be resized. This is mainly limitation of the btrfs tools which currently do not work reliably.

A new btrfs volume can be created with the create command.

The btrfs file system supports subvolume snapshotting, so you can take a snapshot of any btrfs volume in the system with ssm. However btrfs does not distinguish between subvolumes and snapshots, because a snapshot is actually just a subvolume with some blocks shared with a different subvolume. This means, that ssm is not able to directly recognize a btrfs snapshot. Instead, ssm will try to recognize a special name format of the btrfs volume that denotes it is a snapshot. However, if the NAME is specified when creating snapshot which does not match the special pattern, snapshot will not be recognized by the ssm and it will be listed as regular btrfs volume.

A new btrfs snapshot can be created with the snapshot command.

Btrfs does not require a special device to be created on.

Lvm backend

Pools, volumes and snapshots can be created with lvm, which pretty much match the lvm abstraction.

An lvm pool is just a volume group in lvm language. It means that it is grouping devices and new logical volumes can be created out of the lvm pool. The default lvm pool name is lvm_pool.

An lvm pool is created when the create or add commands are used with specified devices and a non existing pool name.

Alternatively a thin pool can be created as a result of using --virtual-size option to create thin volume.

An lvm volume is just a logical volume in lvm language. An lvm volume can be created with the create command.
Lvm volumes can be snapshotted as well. When a snapshot is created from the lvm volume, a new snapshot volume is created, which can be handled as any other lvm volume. Unlike btrfs lvm is able to distinguish snapshot from regular volume, so there is no need for a snapshot name to match special pattern.
Lvm requires a physical device to be created on the device, but with ssm this is transparent for the user.

Crypt backend

The crypt backend in ssm uses cryptsetup and dm-crypt target to manage encrypted volumes. The crypt backend can be used as a regular backend for creating encrypted volumes on top of regular block devices, or even other volumes (lvm or md volumes for example). Or it can be used to create encrypted lvm volumes right away in a single step.

Only volumes can be created with crypt backend. This backend does not support pooling and does not require special devices.

The crypt backend does not support pooling, and it is not possible to create crypt pool or add a device into a pool.
A volume in the crypt backend is the volume created by dm-crypt which represents the data on the original encrypted device in unencrypted form. The crypt backend does not support pooling, so only one device can be used to create crypt volume. It also does not support raid or any device concatenation.

Currently two modes, or extensions are supported: luks and plain. Luks is used by default. For more information about the extensions, please see cryptsetup manual page.

The crypt backend does not support snapshotting, however if the encrypted volume is created on top of an lvm volume, the lvm volume itself can be snapshotted. The snapshot can be then opened by using cryptsetup. It is possible that this might change in the future so that ssm will be able to activate the volume directly without the extra step.
The crypt backend does not require a special device to be created on.

MD backend

MD backend in ssm is currently limited to only gather the information about MD volumes in the system. You can not create or manage MD volumes or pools, but this functionality will be extended in the future.

Multipath backend

Multipath backend in ssm is currently limited to only gather the information about multipath volumes in the system. You can not create or manage multipath volumes or pools, but this functionality will be extended in the future.

EXAMPLES

List system storage information:

# ssm list


List all pools in the system:

# ssm list pools


Create a new 100GB volume with the default lvm backend using /dev/sda and /dev/sdb with xfs file system:

# ssm create --size 100G --fs xfs /dev/sda /dev/sdb


Create a new volume with a btrfs backend using /dev/sda and /dev/sdb and let the volume to be RAID 1:

# ssm -b btrfs create --raid 1 /dev/sda /dev/sdb


Using the lvm backend create a RAID 0 volume with devices /dev/sda and /dev/sdb with 128kB stripe size, ext4 file system and mount it on /home:

# ssm create --raid 0 --stripesize 128k /dev/sda /dev/sdb /home


Create a new thinly provisioned volume with a lvm backend using devices /dev/sda and /dev/sdb using --virtual-size option:

# ssm create --virtual-size 1T /dev/sda /dev/sdb


Create a new thinly provisioned volume with a defined thin pool size and devices /dev/sda and /dev/sdb:

# ssm create --size 50G --virtual-size 1T /dev/sda /dev/sdb


Extend btrfs volume btrfs_pool by 500GB and use /dev/sdc and /dev/sde to cover the resize:

# ssm resize -s +500G btrfs_pool /dev/sdc /dev/sde


Shrink volume /dev/lvm_pool/lvol001 by 1TB:

# ssm resize -s-1t /dev/lvm_pool/lvol001


Remove /dev/sda device from the pool, remove the btrfs_pool pool and also remove the volume /dev/lvm_pool/lvol001:

# ssm remove /dev/sda btrfs_pool /dev/lvm_pool/lvol001


Take a snapshot of the btrfs volume btrfs_pool:my_volume:

# ssm snapshot btrfs_pool:my_volume


Add devices /dev/sda and /dev/sdb into the btrfs_pool pool:

# ssm add -p btrfs_pool /dev/sda /dev/sdb


Mount btrfs subvolume btrfs_pool:vol001 on /mnt/test:

# ssm mount btrfs_pool:vol001 /mnt/test


ENVIRONMENT VARIABLES

Specify which backend will be used by default. This can be overridden by specifying the -b or --backend argument. Currently only lvm and btrfs are supported.
Name of the default lvm pool to be used if the -p or --pool argument is omitted.
Name of the default btrfs pool to be used if the -p or --pool argument is omitted.
When this is set, ssm will filter out all devices, volumes and pools whose name does not start with this prefix. It is used mainly in the ssm test suite to make sure that we do not scramble the local system configuration.

LICENCE

(C)2017 Red Hat, Inc., Jan Tulak <jtulak@redhat.com> (C)2011 Red Hat, Inc., Lukas Czerner <lczerner@redhat.com>

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 2 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program. If not, see <http://www.gnu.org/licenses/>.

REQUIREMENTS

Python 3.6 or higher is required to run this tool. System Storage Manager can only be run as root since most of the commands require root privileges.

There are other requirements listed below, but note that you do not necessarily need all dependencies for all backends. However if some of the tools required by a backend are missing, that backend will not work.

Python modules

  • argparse
  • atexit
  • base64
  • datetime
  • fcntl
  • getpass
  • os
  • pwquality
  • re
  • socket
  • stat
  • struct
  • subprocess
  • sys
  • tempfile
  • termios
  • threading
  • tty

System tools

  • tune2fs
  • fsck.SUPPORTED_FS
  • resize2fs
  • xfs_db
  • xfs_check
  • xfs_growfs
  • mkfs.SUPPORTED_FS
  • which
  • mount
  • blkid
  • wipefs
  • dd

Lvm backend

•
lvm2 binaries

Some distributions (e.g. Debian) have thin provisioning tools for LVM as an optional dependency, while others install it automatically. Thin provisioning without these tools installed is not supported by SSM.

Btrfs backend

•
btrfs progs

Crypt backend

  • dmsetup
  • cryptsetup

Multipath backend

•
multipath

AVAILABILITY

System storage manager is available from http://system-storage-manager.github.io. You can subscribe to storagemanager-devel@lists.sourceforge.net to follow the current development.

AUTHOR

Jan Ťulák <jtulak@redhat.com>, Lukáš Czerner <lczerner@redhat.com>

COPYRIGHT

2015, Red Hat, Inc., Jan Ťulák <jtulak@redhat.com>, Lukáš Czerner <lczerner@redhat.com>

April 7, 2021 1.4