What is system group in linux

cgroups

Control groups (or cgroups as they are commonly known) are a feature provided by the Linux kernel to manage, restrict, and audit groups of processes. Compared to other approaches like the nice(1) command or /etc/security/limits.conf , cgroups are more flexible as they can operate on (sub)sets of processes (possibly with different system users).

Control groups can be accessed with various tools:

• using directives in systemd unit files to specify limits for services and slices;
• by accessing the cgroup filesystem directly;
• via tools like cgcreate , cgexec and cgclassify (part of the libcgroupAUR and libcgroup-gitAUR packages);
• using the «rules engine daemon» to automatically move certain users/groups/commands to groups ( /etc/cgrules.conf and /usr/lib/systemd/system/cgconfig.service ) (part of the libcgroupAUR and libcgroup-gitAUR packages); and
• through other software such as Linux Containers (LXC) virtualization.

cgmanager is deprecated and unsupported as it does not work with systemd versions 232 and above.

For Arch Linux, systemd is the preferred and easiest method of invoking and configuring cgroups as it is a part of the default installation.

Contents

Installing

Make sure you have one of these packages installed for automated cgroup handling:

• systemd — for controlling resources of a systemd service.
• libcgroupAUR , libcgroup-gitAUR — set of standalone tools ( cgcreate , cgclassify , persistence via cgconfig.conf ).

With systemd

Hierarchy

Current cgroup hierarchy can be seen with systemctl status or systemd-cgls command.

Find cgroup of a process

The cgroup name of a process can be found in /proc/PID/cgroup .

For example, the cgroup of the shell:

cgroup resource usage

The systemd-cgtop command can be used to see the resource usage:

Custom cgroups

systemd.slice(5) systemd unit files can be used to define a custom cgroup configuration. They must be placed in a systemd directory, such as /etc/systemd/system/ . The resource control options that can be assigned are documented in systemd.resource-control(5) .

This is an example slice unit that only allows 30% of one CPU to be used:

Remember to run systemctl daemon-reload to pick up any new or changed .slice files.

As service

Service unit file

Resources can be directly specified in service definition or as a Systemd#Drop-in files

Grouping unit under a slice

Service can be specified what slice to run in:

As root

systemd-run can be used to run a command in a specific slice.

—uid=username option can be used to spawn the command as specific user.

The —shell option can be used to spawn a command shell inside the slice.

As unprivileged user

Unprivileged users can divide the resources provided to them into new cgroups, if some conditions are met.

Cgroups v2 must be utilized for a non-root user to be allowed managing cgroup resources.

Controller types

Not all resources can be controlled by user.

Controller	Can be controlled by user	Options
cpu	Requires delegation	CPUAccounting, CPUWeight, CPUQuota, AllowedCPUs, AllowedMemoryNodes
io	Requires delegation	IOWeight, IOReadBandwidthMax, IOWriteBandwidthMax, IODeviceLatencyTargetSec
memory	Yes	MemoryLow, MemoryHigh, MemoryMax, MemorySwapMax
pids	Yes	TasksMax
rdma	No	?
eBPF	No	IPAddressDeny, DeviceAllow, DevicePolicy

User delegation

For user to control cpu and io resources, the resources need to be delegated. This can be done by creating a unit overload.

For example if your user id is 1000:

Reboot and verify that the slice your user session is under has cpu and io controller:

User-defined slices

The user slice files can be placed in

To run the command under certain slice:

You can also run your login shell inside the slice:

Run-time adjustment

cgroups resources can be adjusted at run-time using systemctl set-property command. Option syntax is the same as in systemd.resource-control(5) .

For example, cutting off internet access for all user sessions:

With libcgroup

You can enable the cgconfig service with systemd. This allows you to track any errors in cgconfig.conf more easily.

Ad-hoc groups

One of the powers of cgroups is that you can create «ad-hoc» groups on the fly. You can even grant the privileges to create custom groups to regular users. groupname is the cgroup name:

Now all the tunables in the group groupname are writable by your user:

Cgroups are hierarchical, so you can create as many subgroups as you like. If a normal user wants to run a bash shell under a new subgroup called foo :

To make sure (only meaningful for legacy (v1) cgroups):

A new subdirectory was created for this group. To limit the memory usage of all processes in this group to 10 MB, run the following:

Note that the memory limit applies to RAM use only — once tasks hit this limit, they will begin to swap. But it won’t affect the performance of other processes significantly.

Similarly you can change the CPU priority («shares») of this group. By default all groups have 1024 shares. A group with 100 shares will get a

10% portion of the CPU time:

You can find more tunables or statistics by listing the cgroup directory.

You can also change the cgroup of already running processes. To move all ‘bash’ commands to this group:

Persistent group configuration

If you want your cgroups to be created at boot, you can define them in /etc/cgconfig.conf instead. For example, the «groupname» has a permission for $USER and users of group $GROUP to manage limits and add tasks. A subgroup «groupname/foo» group definitions would look like this:

This is equivalent to these shell commands:

With the cgroup virtual filesystem

This article or section needs expansion.

Starting with systemd 232, the cgm method described in the next section, this section will instead describe a manual method to limit memory usage.

Create a new cgroup named groupname:

Example: set the maximum memory limit to 100MB:

Move a process to the cgroup (note: only one PID can be written at a time, repeat this for each process that must be moved):

Examples

Matlab

Doing large calculations in MATLAB can crash your system, because Matlab does not have any protection against taking all your machine’s memory or CPU. The following examples show a cgroup that constrains Matlab to first 6 CPU cores and 5 GB of memory.

With systemd

Launch Matlab like this (be sure to use the right path):

With libcgroup

Change username to the user Matlab is run as.

You can also restrict the CPU share with the cpu constraint.

Launch Matlab like this (be sure to use the right path):

Documentation

• For information on controllers and what certain switches and tunables mean, refer to kernel’s documentation v1 or v2 (or install linux-docs and see /usr/src/linux/Documentation/cgroup )
• A detailed and complete Resource Management Guide can be found in the fedora project documentation.

For commands and configuration files, see relevant man pages, e.g. man cgcreate or man cgrules.conf

Tips and tricks

Enable cgroup v1

Cgroup v2 is now enabled by default. If you want to switch to cgroup v1 instead, you need to set the following kernel parameter:

Источник

Users and groups

Users and groups are used on GNU/Linux for access control—that is, to control access to the system’s files, directories, and peripherals. Linux offers relatively simple/coarse access control mechanisms by default. For more advanced options, see ACL, Capabilities and PAM#Configuration How-Tos.

Contents

Overview

A user is anyone who uses a computer. In this case, we are describing the names which represent those users. It may be Mary or Bill, and they may use the names Dragonlady or Pirate in place of their real name. All that matters is that the computer has a name for each account it creates, and it is this name by which a person gains access to use the computer. Some system services also run using restricted or privileged user accounts.

Managing users is done for the purpose of security by limiting access in certain specific ways. The superuser (root) has complete access to the operating system and its configuration; it is intended for administrative use only. Unprivileged users can use the su and sudo programs for controlled privilege elevation.

Any individual may have more than one account as long as they use a different name for each account they create. Further, there are some reserved names which may not be used such as «root».

Users may be grouped together into a «group», and users may be added to an existing group to utilize the privileged access it grants.

Permissions and ownership

The UNIX operating system crystallizes a couple of unifying ideas and concepts that shaped its design, user interface, culture and evolution. One of the most important of these is probably the mantra: «everything is a file,» widely regarded as one of the defining points of UNIX. This key design principle consists of providing a unified paradigm for accessing a wide range of input/output resources: documents, directories, hard-drives, CD-ROMs, modems, keyboards, printers, monitors, terminals and even some inter-process and network communications. The trick is to provide a common abstraction for all of these resources, each of which the UNIX fathers called a «file.» Since every «file» is exposed through the same API, you can use the same set of basic commands to read/write to a disk, keyboard, document or network device.

A fundamental and very powerful, consistent abstraction provided in UNIX and compatible operating systems is the file abstraction. Many OS services and device interfaces are implemented to provide a file or file system metaphor to applications. This enables new uses for, and greatly increases the power of, existing applications — simple tools designed with specific uses in mind can, with UNIX file abstractions, be used in novel ways. A simple tool, such as cat, designed to read one or more files and output the contents to standard output, can be used to read from I/O devices through special device files, typically found under the /dev directory. On many systems, audio recording and playback can be done simply with the commands, » cat /dev/audio > myfile » and » cat myfile > /dev/audio ,» respectively.

Every file on a GNU/Linux system is owned by a user and a group. In addition, there are three types of access permissions: read, write, and execute. Different access permissions can be applied to a file’s owning user, owning group, and others (those without ownership). One can determine a file’s owners and permissions by viewing the long listing format of the ls command:

The first column displays the file’s permissions (for example, the file initramfs-linux.img has permissions -rw-r—r— ). The third and fourth columns display the file’s owning user and group, respectively. In this example, all files are owned by the root user and the root group.

In this example, the sf_Shared directory is owned by the root user and the vboxsf group. It is also possible to determine a file’s owners and permissions using the stat command:

Access permissions are displayed in three groups of characters, representing the permissions of the owning user, owning group, and others, respectively. For example, the characters -rw-r—r— indicate that the file’s owner has read and write permission, but not execute ( rw- ), whilst users belonging to the owning group and other users have only read permission ( r— and r— ). Meanwhile, the characters drwxrwx— indicate that the file’s owner and users belonging to the owning group all have read, write, and execute permissions ( rwx and rwx ), whilst other users are denied access ( — ). The first character represents the file’s type.

List files owned by a user or group with the find utility:

A file’s owning user and group can be changed with the chown (change owner) command. A file’s access permissions can be changed with the chmod (change mode) command.

Shadow

The user, group and password management tools on Arch Linux come from the shadow package, which is a dependency of the base meta package.

File list

File	Purpose
/etc/shadow	Secure user account information
/etc/passwd	User account information
/etc/gshadow	Contains the shadowed information for group accounts
/etc/group	Defines the groups to which users belong

User management

To list users currently logged on the system, the who command can be used. To list all existing user accounts including their properties stored in the user database, run passwd -Sa as root. See passwd(1) for the description of the output format.

To add a new user, use the useradd command:

-m / —create-home the user’s home directory is created as /home/username . The directory is populated by the files in the skeleton directory. The created files are owned by the new user. -G / —groups a comma separated list of supplementary groups which the user is also a member of. The default is for the user to belong only to the initial group. -s / —shell a path to the user’s login shell. Ensure the chosen shell is installed if choosing something other than Bash.

If an initial login group is specified by name or number, it must refer to an already existing group. If not specified, the behaviour of useradd will depend on the USERGROUPS_ENAB variable contained in /etc/login.defs . The default behaviour ( USERGROUPS_ENAB yes ) is to create a group with the same name as the username.

When the login shell is intended to be non-functional, for example when the user account is created for a specific service, /usr/bin/nologin may be specified in place of a regular shell to politely refuse a login (see nologin(8) ).

See useradd(8) for other supported options.

Example adding a user

To add a new user named archie , creating its home directory and otherwise using all the defaults in terms of groups, folder names, shell used and various other parameters:

Although it is not required to protect the newly created user archie with a password, it is highly recommended to do so:

The above useradd command will also automatically create a group called archie and makes this the default group for the user archie . Making each user have their own group (with the group name same as the user name) is the preferred way to add users.

You could also make the default group something else using the -g option, but note that, in multi-user systems, using a single default group (e.g. users ) for every user is not recommended. The reason is that typically, the method for facilitating shared write access for specific groups of users is setting user umask value to 002 , which means that the default group will by default always have write access to any file you create. See also User Private Groups. If a user must be a member of a specific group specify that group as a supplementary group when creating the user.

In the recommended scenario, where the default group has the same name as the user name, all files are by default writeable only for the user who created them. To allow write access to a specific group, shared files/folders can be made writeable by default for everyone in this group and the owning group can be automatically fixed to the group which owns the parent directory by setting the setgid bit on this directory:

Otherwise the file creator’s default group (usually the same as the user name) is used.

If a GID change is required temporarily you can also use the newgrp command to change the user’s default GID to another GID at runtime. For example, after executing newgrp groupname files created by the user will be associated with the groupname GID, without requiring a re-login. To change back to the default GID, execute newgrp without a groupname.

Example adding a system user

System users can be used to run processes/daemons under a different user, protecting (e.g. with chown) files and/or directories and more examples of computer hardening.

With the following command a system user without shell access and without a home directory is created (optionally append the -U parameter to create a group with the same name as the user, and add the user to this group):

If the system user requires a specific user and group ID, specify them with the -u / —uid and -g / —gid options when creating the user:

Change a user’s login name or home directory

To change a user’s home directory:

The -m option also automatically creates the new directory and moves the content there.

Make sure there is no trailing / on /my/old/home .

To change a user’s login name:

Changing a username is safe and easy when done properly, just use the usermod command. If the user is associated to a group with the same name, you can rename this with the groupmod command.

Alternatively, the /etc/passwd file can be edited directly, see #User database for an introduction to its format.

Also keep in mind the following notes:

• If you are using sudo make sure you update your /etc/sudoers to reflect the new username(s) (via the visudo command as root).
• Personal crontabs need to be adjusted by renaming the user’s file in /var/spool/cron from the old to the new name, and then opening crontab -e to change any relevant paths and have it adjust the file permissions accordingly.
• Wine’s personal folders/files’ contents in

/.local/share/applications/wine/Programs and possibly more need to be manually renamed/edited.

Certain Thunderbird addons, like Enigmail, may need to be reinstalled.

Anything on your system (desktop shortcuts, shell scripts, etc.) that uses an absolute path to your home dir (i.e. /home/oldname ) will need to be changed to reflect your new name. To avoid these problems in shell scripts, simply use the

or $HOME variables for home directories.

Also do not forget to edit accordingly the configuration files in /etc/ that relies on your absolute path (e.g. Samba, CUPS, so on). A nice way to learn what files you need to update involves using the grep command this way: grep -r old_user *

Other examples of user management

To enter user information for the GECOS comment (e.g. the full user name), type:

(this way chfn runs in interactive mode).

Alternatively the GECOS comment can be set more liberally with:

To mark a user’s password as expired, requiring them to create a new password the first time they log in, type:

User accounts may be deleted with the userdel command:

The -r option specifies that the user’s home directory and mail spool should also be deleted.

To change the user’s login shell:

User database

Local user information is stored in the plain-text /etc/passwd file: each of its lines represents a user account, and has seven fields delimited by colons.

• account is the user name. This field can not be blank. Standard *NIX naming rules apply.
• password is the user password.

Broken down, this means: user jack , whose password is in /etc/shadow , whose UID is 1001 and whose primary group is 1003. Jack Smith is his full name and there is a comment associated to his account; his home directory is /home/jack and he is using Bash.

The pwck command can be used to verify the integrity of the user database. It can sort the user list by GID at the same time, which can be helpful for comparison:

Automatic integrity checks

Instead of running pwck / grpck manually, the systemd timer shadow.timer , which is part of, and is enabled by, installation of the shadow package, will start shadow.service daily. shadow.service will run pwck(8) and grpck(8) to verify the integrity of both password and group files.

If discrepancies are reported, group can be edited with the vigr(8) command and users with vipw(8) . This provides an extra margin of protection in that these commands lock the databases for editing. Note that the default text editor is vi, but an alternative editor will be used if the EDITOR environment variable is set, then that editor will be used instead.

Group management

/etc/group is the file that defines the groups on the system (see group(5) for details). There is also its companion gshadow which is rarely used. Its details are at gshadow(5) .

Display group membership with the groups command:

If user is omitted, the current user’s group names are displayed.

The id command provides additional detail, such as the user’s UID and associated GIDs:

To list all groups on the system:

Create new groups with the groupadd command:

Add users to a group with the gpasswd command (see FS#58262 regarding errors):

Alternatively, add a user to additional groups with usermod (replace additional_groups with a comma-separated list):

Modify an existing group with the groupmod command, e.g. to rename the old_group group to new_group :

To delete existing groups:

To remove users from a group:

The grpck command can be used to verify the integrity of the system’s group files.

Group list

This section explains the purpose of the essential groups from the filesystem package. There are many other groups, which will be created with correct GID when the relevant package is installed. See the main page for the software for details.

User groups

Non-root workstation/desktop users often need to be added to some of following groups to allow access to hardware peripherals and facilitate system administration:

Group	Affected files	Purpose
adm	Administration group, commonly used to give read access to protected logs. It has full read access to journal files.
ftp	/srv/ftp/	Access to files served by FTP servers.
games	/var/games	Access to some game software.
http	/srv/http/	Access to files served by HTTP servers.
log	Access to log files in /var/log/ created by syslog-ng.
rfkill	/dev/rfkill	Right to control wireless devices power state (used by rfkill).
sys	Right to administer printers in CUPS.
systemd-journal	/var/log/journal/*	Can be used to provide read-only access to the systemd logs, as an alternative to adm and wheel [1]. Otherwise, only user generated messages are displayed.
uucp	/dev/ttyS3+ , /dev/tts/9+ , /dev/ttyUSB3+ , /dev/ttyACM2+ , /dev/rfcomm5+	RS-232 serial ports and devices connected to them.
wheel	Administration group, commonly used to give privileges to perform administrative actions. It has full read access to journal files and the right to administer printers in CUPS. Can also be used to give access to the sudo and su utilities (neither uses it by default).

System groups

The following groups are used for system purposes, an assignment to users is only required for dedicated purposes:

Group	Affected files	Purpose
dbus	used internally by dbus
kmem	/dev/port , /dev/mem , /dev/kmem
locate	/usr/bin/locate , /var/lib/locate , /var/lib/mlocate , /var/lib/slocate	See Locate.
lp	/dev/lp7* , /dev/parport3*	Access to parallel port devices (printers and others).
mail	/usr/bin/mail
nobody	Unprivileged group.
proc	/proc/pid/	A group authorized to learn processes information otherwise prohibited by hidepid= mount option of the proc file system. The group must be explicitly set with the gid= mount option.
root	/*	Complete system administration and control (root, admin).
smmsp	sendmail group.
tty	/dev/tty , /dev/vcc , /dev/vc , /dev/ptmx
utmp	/run/utmp , /var/log/btmp , /var/log/wtmp

Pre-systemd groups

Before arch migrated to systemd, users had to be manually added to these groups in order to be able to access the corresponding devices. This way has been deprecated in favour of udev marking the devices with a uaccess tag and logind assigning the permissions to users dynamically via ACLs according to which session is currently active. Note that the session must not be broken for this to work (see General troubleshooting#Session permissions to check it).

There are some notable exceptions which require adding a user to some of these groups: for example if you want to allow users to access the device even when they are not logged in. However, note that adding users to the groups can even cause some functionality to break (for example, the audio group will break fast user switching and allows applications to block software mixing).

Group	Affected files	Purpose
audio	/dev/audio , /dev/snd/* , /dev/rtc0	Direct access to sound hardware, for all sessions. It is still required to make ALSA and OSS work in remote sessions, see ALSA#User privileges. Also used in JACK to give users realtime processing permissions.
disk	/dev/sd[a-zA-Z]*2*	Access to block devices not affected by other groups such as optical , floppy , and storage .
floppy	/dev/fd1*	Access to floppy drives.
input	/dev/input/event7* , /dev/input/mouse5*	Access to input devices. Introduced in systemd 215 [2].
kvm	/dev/kvm	Access to virtual machines using KVM.
optical	/dev/sr6 , /dev/sg3	Access to optical devices such as CD and DVD drives.
scanner	/var/lock/sane	Access to scanner hardware.
storage	/dev/st6*[lma]* , /dev/nst3*[lma]*	Used to gain access to removable drives such as USB hard drives, flash/jump drives, MP3 players; enables the user to mount storage devices.[3] Now solely for direct access to tapes if no custom udev rules is involved.[4][5][6][7]
video	/dev/fb/0 , /dev/misc/agpgart	Access to video capture devices, 2D/3D hardware acceleration, framebuffer (X can be used without belonging to this group).

Unused groups

The following groups are currently not used for any purpose:

Group	Affected files	Purpose
bin	none	Historical
daemon
lock	Used for lockfile access. Required by e.g. gnokii AUR .
mem
network	Unused by default. Can be used e.g. for granting access to NetworkManager (see NetworkManager#Set up PolicyKit permissions).
power
uuidd
users	The primary group for users when user private groups are not used (generally not recommended), e.g. when creating users with USERGROUPS_ENAB no in /etc/login.defs or the -N / —no-user-group option of useradd.

Other tools related to these databases

This article or section is a candidate for merging with #Shadow.

The factual accuracy of this article or section is disputed.

getent(1) can be used to read a particular record.

As warned in #User database, using specific utilities such as passwd and chfn , is a better way to change the databases. Nevertheless, there are times when editing them directly is looked after. For those times, vipw , vigr are provided. It is strongly recommended to use these tailored editors over using a general text editor as they lock the databases against concurrent editing. They also help prevent invalid entries and/or syntax errors. Note that Arch Linux prefers usage of specific tools, such as chage, for modifying the shadow database over using vipw -s and vigr -s from util-linux . See also FS#31414.

Источник