1podman-create(1)() podman-create(1)()
2
6 podman-create - Create a new container
7
10 podman create [options] image [command [arg ...]]
11 podman container create [options] image [command [arg ...]]
14
17 Creates a writable container layer over the specified image and pre‐
18 pares it for running the specified command. The container ID is then
19 printed to STDOUT. This is similar to podman run -d except the con‐
20 tainer is never started. You can then use the podman start container
21 command to start the container at any point.
22 The initial status of the container created with podman create is 'cre‐
25 ated'.
26 Default settings for flags are defined in containers.conf. Most set‐
29 tings for remote connections use the server's containers.conf, except
30 when documented in man pages.
31
34 The image is specified using transport:path format. If no transport is
35 specified, the docker (container registry) transport will be used by
36 default. For remote Podman, docker is the only allowed transport.
37 dir:path
40 An existing local directory path storing the manifest, layer tarballs
41 and signatures as individual files. This is a non-standardized format,
42 primarily useful for debugging or noninvasive container inspection.
43 $ podman save --format docker-dir fedora -o /tmp/fedora
46 $ podman create dir:/tmp/fedora echo hello
47 docker://docker-reference (Default)
51 An image reference stored in a remote container image registry. Ex‐
52 ample: "quay.io/podman/stable:latest". The reference can include a
53 path to a specific registry; if it does not, the registries listed in
54 registries.conf will be queried to find a matching image. By default,
55 credentials from podman login (stored at $XDG_RUNTIME_DIR/contain‐
56 ers/auth.json by default) will be used to authenticate; otherwise it
57 falls back to using credentials in $HOME/.docker/config.json.
58 $ podman create registry.fedoraproject.org/fedora:latest echo hello
61 docker-archive:path[:docker-reference] An image stored in the docker
65 save formatted file. docker-reference is only used when creating such a
66 file, and it must not contain a digest.
67 $ podman save --format docker-archive fedora -o /tmp/fedora
70 $ podman create docker-archive:/tmp/fedora echo hello
71 docker-daemon:docker-reference
75 An image in docker-reference format stored in the docker daemon in‐
76 ternal storage. The docker-reference can also be an image ID
77 (docker-daemon:algo:digest).
78 $ sudo docker pull fedora
81 $ sudo podman create docker-daemon:docker.io/library/fedora echo hello
82 oci-archive:path:tag
86 An image in a directory compliant with the "Open Container Image Lay‐
87 out Specification" at the specified path and specified with a tag.
88 $ podman save --format oci-archive fedora -o /tmp/fedora
91 $ podman create oci-archive:/tmp/fedora echo hello
92
96 --add-host=host
97 Add a custom host-to-IP mapping (host:ip)
98 Add a line to /etc/hosts. The format is hostname:ip. The --add-host op‐
101 tion can be set multiple times.
102 --annotation=key=value
105 Add an annotation to the container. The format is key=value. The --an‐
106 notation option can be set multiple times.
107 --arch=ARCH
110 Override the architecture, defaults to hosts, of the image to be
111 pulled. For example, arm.
112 --attach, -a=location
115 Attach to STDIN, STDOUT or STDERR.
116 In foreground mode (the default when -d is not specified), podman run
119 can start the process in the container and attach the console to the
120 process's standard input, output, and standard error. It can even pre‐
121 tend to be a TTY (this is what most command line executables expect)
122 and pass along signals. The -a option can be set for each of stdin,
123 stdout, and stderr.
124 --authfile=path
127 Path of the authentication file. Default is ${XDG_RUNTIME_DIR}/contain‐
128 ers/auth.json
129 Note: You can also override the default path of the authentication file
132 by setting the REGISTRY_AUTH_FILE environment variable. export REG‐
133 ISTRY_AUTH_FILE=path
134 --blkio-weight=weight
137 Block IO weight (relative weight) accepts a weight value between 10 and
138 1000.
139 --blkio-weight-device=weight
142 Block IO weight (relative device weight, format: DEVICE_NAME:WEIGHT).
143 --cap-add=capability
146 Add Linux capabilities
147 --cap-drop=capability
150 Drop Linux capabilities
151 --cgroupns=mode
154 Set the cgroup namespace mode for the container.
155 host: use the host's cgroup namespace inside the container.
156 container:<NAME|ID>: join the namespace of the specified container.
157 ns:<PATH>: join the namespace at the specified path.
158 private: create a new cgroup namespace.
159 If the host uses cgroups v1, the default is set to host. On cgroups v2
162 the default is private.
163 --cgroups=mode
166 Determines whether the container will create CGroups. Valid values are
167 enabled, disabled, no-conmon, split, which the default being enabled.
168 The enabled option will create a new cgroup under the cgroup-parent.
171 The disabled option will force the container to not create CGroups, and
172 thus conflicts with CGroup options (--cgroupns and --cgroup-parent).
173 The no-conmon option disables a new CGroup only for the conmon process.
174 The split option splits the current cgroup in two sub-cgroups: one for
175 conmon and one for the container payload. It is not possible to set
176 --cgroup-parent with split.
177 --cgroup-parent=path
180 Path to cgroups under which the cgroup for the container will be cre‐
181 ated. If the path is not absolute, the path is considered to be rela‐
182 tive to the cgroups path of the init process. Cgroups will be created
183 if they do not already exist.
184 --cgroup-conf=KEY=VALUE
187 When running on cgroup v2, specify the cgroup file to write to and its
188 value. For example --cgroup-conf=memory.high=1073741824 sets the mem‐
189 ory.high limit to 1GB.
190 --cidfile=id
193 Write the container ID to the file
194 --conmon-pidfile=path
197 Write the pid of the conmon process to a file. conmon runs in a sepa‐
198 rate process than Podman, so this is necessary when using systemd to
199 restart Podman containers. (This option is not available with the re‐
200 mote Podman client)
201 --cpu-period=limit
204 Set the CPU period for the Completely Fair Scheduler (CFS), which is a
205 duration in microseconds. Once the container's CPU quota is used up, it
206 will not be scheduled to run until the current period ends. Defaults to
207 100000 microseconds.
208 On some systems, changing the CPU limits may not be allowed for
211 non-root users. For more details, see https://github.com/contain‐
212 ers/podman/blob/master/troubleshooting.md#26-running-contain‐
213 ers-with-cpu-limits-fails-with-a-permissions-error
214 --cpu-quota=limit
217 Limit the CPU Completely Fair Scheduler (CFS) quota.
218 Limit the container's CPU usage. By default, containers run with the
221 full CPU resource. The limit is a number in microseconds. If you pro‐
222 vide a number, the container will be allowed to use that much CPU time
223 until the CPU period ends (controllable via --cpu-period).
224 On some systems, changing the CPU limits may not be allowed for
227 non-root users. For more details, see https://github.com/contain‐
228 ers/podman/blob/master/troubleshooting.md#26-running-contain‐
229 ers-with-cpu-limits-fails-with-a-permissions-error
230 --cpu-rt-period=microseconds
233 Limit the CPU real-time period in microseconds
234 Limit the container's Real Time CPU usage. This flag tell the kernel to
237 restrict the container's Real Time CPU usage to the period you specify.
238 --cpu-rt-runtime=microseconds
241 Limit the CPU real-time runtime in microseconds
242 Limit the containers Real Time CPU usage. This flag tells the kernel to
245 limit the amount of time in a given CPU period Real Time tasks may con‐
246 sume. Ex: Period of 1,000,000us and Runtime of 950,000us means that
247 this container could consume 95% of available CPU and leave the remain‐
248 ing 5% to normal priority tasks.
249 The sum of all runtimes across containers cannot exceed the amount al‐
252 lotted to the parent cgroup.
253 --cpu-shares=shares
256 CPU shares (relative weight)
257 By default, all containers get the same proportion of CPU cycles. This
260 proportion can be modified by changing the container's CPU share
261 weighting relative to the weighting of all other running containers.
262 To modify the proportion from the default of 1024, use the --cpu-shares
265 flag to set the weighting to 2 or higher.
266 The proportion will only apply when CPU-intensive processes are run‐
269 ning. When tasks in one container are idle, other containers can use
270 the left-over CPU time. The actual amount of CPU time will vary depend‐
271 ing on the number of containers running on the system.
272 For example, consider three containers, one has a cpu-share of 1024 and
275 two others have a cpu-share setting of 512. When processes in all three
276 containers attempt to use 100% of CPU, the first container would re‐
277 ceive 50% of the total CPU time. If you add a fourth container with a
278 cpu-share of 1024, the first container only gets 33% of the CPU. The
279 remaining containers receive 16.5%, 16.5% and 33% of the CPU.
280 On a multi-core system, the shares of CPU time are distributed over all
283 CPU cores. Even if a container is limited to less than 100% of CPU
284 time, it can use 100% of each individual CPU core.
285 For example, consider a system with more than three cores. If you start
288 one container {C0} with -c=512 running one process, and another con‐
289 tainer {C1} with -c=1024 running two processes, this can result in the
290 following division of CPU shares:
291 PID container CPU CPU share 100 {C0} 0 100% of
294 CPU0 101 {C1} 1 100% of CPU1 102
295 {C1} 2 100% of CPU2
296 --cpus=number
299 Number of CPUs. The default is 0.0 which means no limit. This is short‐
300 hand for --cpu-period and --cpu-quota, so you may only set either
301 --cpus or --cpu-period and --cpu-quota.
304 On some systems, changing the CPU limits may not be allowed for
305 non-root users. For more details, see https://github.com/contain‐
306 ers/podman/blob/master/troubleshooting.md#26-running-contain‐
307 ers-with-cpu-limits-fails-with-a-permissions-error
308 --cpuset-cpus=cpus
311 CPUs in which to allow execution (0-3, 0,1)
312 --cpuset-mems=nodes
315 Memory nodes (MEMs) in which to allow execution (0-3, 0,1). Only effec‐
316 tive on NUMA systems.
317 If you have four memory nodes on your system (0-3), use
320 --cpuset-mems=0,1 then processes in your container will only use memory
321 from the first two memory nodes.
322 --device=host-device[:container-device][:permissions]
325 Add a host device to the container. Optional permissions parameter can
326 be used to specify device permissions, it is combination of r for read,
327 w for write, and m for mknod(2).
328 Example: --device=/dev/sdc:/dev/xvdc:rwm.
331 Note: if _hostdevice is a symbolic link then it will be resolved first.
334 The container will only store the major and minor numbers of the host
335 device.
336 Note: if the user only has access rights via a group, accessing the de‐
339 vice from inside a rootless container will fail. Use the --group-add
340 keep-groups flag to pass the user's supplementary group access into the
341 container.
342 Podman may load kernel modules required for using the specified device.
345 The devices that podman will load modules when necessary are:
346 /dev/fuse.
347 --device-cgroup-rule="type major:minor mode"
350 Add a rule to the cgroup allowed devices list. The rule is expected to
351 be in the format specified in the Linux kernel documentation (Documen‐
352 tation/cgroup-v1/devices.txt):
353 - type: a (all), c (char), or b (block);
354 - major and minor: either a number, or * for all;
355 - mode: a composition of r (read), w (write), and m (mknod(2)).
356 --device-read-bps=path
359 Limit read rate (bytes per second) from a device (e.g. --de‐
360 vice-read-bps=/dev/sda:1mb)
361 --device-read-iops=path
364 Limit read rate (IO per second) from a device (e.g. --de‐
365 vice-read-iops=/dev/sda:1000)
366 --device-write-bps=path
369 Limit write rate (bytes per second) to a device (e.g. --de‐
370 vice-write-bps=/dev/sda:1mb)
371 --device-write-iops=path
374 Limit write rate (IO per second) to a device (e.g. --de‐
375 vice-write-iops=/dev/sda:1000)
376 --disable-content-trust
379 This is a Docker specific option to disable image verification to a
380 Docker registry and is not supported by Podman. This flag is a NOOP and
381 provided solely for scripting compatibility.
382 --dns=dns
385 Set custom DNS servers. Invalid if using --dns and --network that is
386 set to 'none' or container:<name|id>.
387 This option can be used to override the DNS configuration passed to the
390 container. Typically this is necessary when the host DNS configuration
391 is invalid for the container (e.g., 127.0.0.1). When this is the case
392 the --dns flags is necessary for every run.
393 The special value none can be specified to disable creation of /etc/re‐
396 solv.conf in the container by Podman. The /etc/resolv.conf file in the
397 image will be used without changes.
398 --dns-opt=option
401 Set custom DNS options. Invalid if using --dns-opt and --network that
402 is set to 'none' or container:<name|id>.
403 --dns-search=domain
406 Set custom DNS search domains. Invalid if using --dns-search and --net‐
407 work that is set to 'none' or container:<name|id>. (Use --dns-search=.
408 if you don't wish to set the search domain)
409 --entrypoint="command" | '["command", arg1 , ...]'
412 Overwrite the default ENTRYPOINT of the image
413 This option allows you to overwrite the default entrypoint of the im‐
416 age. The ENTRYPOINT of an image is similar to a COMMAND because it
417 specifies what executable to run when the container starts, but it is
418 (purposely) more difficult to override. The ENTRYPOINT gives a con‐
419 tainer its default nature or behavior, so that when you set an ENTRY‐
420 POINT you can run the container as if it were that binary, complete
421 with default options, and you can pass in more options via the COMMAND.
422 But, sometimes an operator may want to run something else inside the
423 container, so you can override the default ENTRYPOINT at runtime by us‐
424 ing a --entrypoint and a string to specify the new ENTRYPOINT.
425 You need to specify multi option commands in the form of a json string.
428 --env, -e=env
431 Set environment variables
432 This option allows arbitrary environment variables that are available
435 for the process to be launched inside of the container. If an environ‐
436 ment variable is specified without a value, Podman will check the host
437 environment for a value and set the variable only if it is set on the
438 host. If an environment variable ending in * is specified, Podman will
439 search the host environment for variables starting with the prefix and
440 will add those variables to the container. If an environment variable
441 with a trailing ***** is specified, then a value must be supplied.
442 See Environment ⟨#environment⟩ note below for precedence and examples.
445 --env-host=true|false
448 Use host environment inside of the container. See Environment note be‐
449 low for precedence. (This option is not available with the remote Pod‐
450 man client)
451 --env-file=file
454 Read in a line delimited file of environment variables. See Environment
455 note below for precedence.
456 --expose=port
459 Expose a port, or a range of ports (e.g. --expose=3300-3310) to set up
460 port redirection on the host system.
461 --gidmap=container_gid:host_gid:amount
464 GID map for the user namespace. Using this flag will run the container
465 with user namespace enabled. It conflicts with the --userns and --sub‐
466 gidname flags.
467 The following example maps uids 0-2000 in the container to the uids
470 30000-31999 on the host and gids 0-2000 in the container to the gids
471 30000-31999 on the host. --gidmap=0:30000:2000
472 --group-add=group|keep-groups
475 Add additional groups to assign to primary user running within the con‐
476 tainer process.
477 • keep-groups is a special flag that tells Podman to keep the
480 supplementary group access.
481 Allows container to use the user's supplementary group access. If file
485 systems or devices are only accessible by the rootless user's group,
486 this flag tells the OCI runtime to pass the group access into the con‐
487 tainer. Currently only available with the crun OCI runtime. Note:
488 keep-groups is exclusive, you cannot add any other groups with this
489 flag. (Not available for remote commands)
490 --health-cmd="command" | '["command", arg1 , ...]'
493 Set or alter a healthcheck command for a container. The command is a
494 command to be executed inside your container that determines your con‐
495 tainer health. The command is required for other healthcheck options to
496 be applied. A value of none disables existing healthchecks.
497 Multiple options can be passed in the form of a JSON array; otherwise,
500 the command will be interpreted as an argument to /bin/sh -c.
501 --health-interval=interval
504 Set an interval for the healthchecks (a value of disable results in no
505 automatic timer setup) (default "30s")
506 --health-retries=retries
509 The number of retries allowed before a healthcheck is considered to be
510 unhealthy. The default value is 3.
511 --health-start-period=period
514 The initialization time needed for a container to bootstrap. The value
515 can be expressed in time format like 2m3s. The default value is 0s
516 --health-timeout=timeout
519 The maximum time allowed to complete the healthcheck before an interval
520 is considered failed. Like start-period, the value can be expressed in
521 a time format such as 1m22s. The default value is 30s.
522 --hostname=name, -h
525 Container host name
526 Sets the container host name that is available inside the container.
529 Can only be used with a private UTS namespace --uts=private (default).
530 If --pod is specified and the pod shares the UTS namespace (default)
531 the pod's hostname will be used.
532 --help
535 Print usage statement
536 --http-proxy=true|false
539 By default proxy environment variables are passed into the container if
540 set for the Podman process. This can be disabled by setting the
541 --http-proxy option to false. The environment variables passed in in‐
542 clude http_proxy, https_proxy, ftp_proxy, no_proxy, and also the upper
543 case versions of those. This option is only needed when the host system
544 must use a proxy but the container should not use any proxy. Proxy en‐
545 vironment variables specified for the container in any other way will
546 override the values that would have been passed through from the host.
547 (Other ways to specify the proxy for the container include passing the
548 values with the --env flag, or hard coding the proxy environment at
549 container build time.) (This option is not available with the remote
550 Podman client)
551 For example, to disable passing these environment variables from host
554 to container:
555 --http-proxy=false
558 Defaults to true
561 --image-volume, builtin-volume=bind|tmpfs|ignore
564 Tells Podman how to handle the builtin image volumes. Default is bind.
565 • bind: An anonymous named volume will be created and mounted
568 into the container.
569 • tmpfs: The volume is mounted onto the container as a tmpfs,
571 which allows the users to create content that disappears when
572 the container is stopped.
573 • ignore: All volumes are just ignored and no action is taken.
575 --init
579 Run an init inside the container that forwards signals and reaps pro‐
580 cesses.
581 --init-path=path
584 Path to the container-init binary.
585 --interactive, -i=true|false
588 Keep STDIN open even if not attached. The default is false.
589 --ip6=ip
592 Not implemented
593 --ip=ip
596 Specify a static IP address for the container, for example
597 10.88.64.128. This option can only be used if the container is joined
598 to only a single network - i.e., --network=_network-name_ is used at
599 most once - and if the container is not joining another container's
600 network namespace via --network=container:_id_. The address must be
601 within the CNI network's IP address pool (default 10.88.0.0/16).
602 --ipc=ipc
605 Default is to create a private IPC namespace (POSIX SysV IPC) for the
606 container container:<name|id>: reuses another container
607 shared memory, semaphores and message queues host: use the
608 host shared memory,semaphores and message queues inside the container.
609 Note: the host mode gives the container full access to local shared
610 memory and is therefore considered insecure. ns:<path> path
611 to an IPC namespace to join.
612 --kernel-memory=number[unit]
615 Kernel memory limit (format: <number>[<unit>], where unit = b (bytes),
616 k (kilobytes), m (megabytes), or g (gigabytes))
617 Constrains the kernel memory available to a container. If a limit of 0
620 is specified (not using --kernel-memory), the container's kernel memory
621 is not limited. If you specify a limit, it may be rounded up to a mul‐
622 tiple of the operating system's page size and the value can be very
623 large, millions of trillions.
624 --label, -l=label
627 Add metadata to a container (e.g., --label com.example.key=value)
628 --label-file=file
631 Read in a line delimited file of labels
632 --link-local-ip=ip
635 Not implemented
636 --log-driver="k8s-file"
639 Logging driver for the container. Currently available options are
640 k8s-file, journald, and none, with json-file aliased to k8s-file for
641 scripting compatibility.
642 --log-opt=name=value
645 Set custom logging configuration. The following *name*s are supported:
646 • path: specify a path to the log file (e.g. --log-opt
649 path=/var/log/container/mycontainer.json);
650 • max-size: specify a max size of the log file (e.g. --log-opt
652 max-size=10mb);
653 • tag: specify a custom log tag for the container (e.g.
655 --log-opt tag="{{.ImageName}}".
656 It supports the same keys as podman inspect --format.
660 This option is currently supported only by the journald log driver.
663 --log-opt tag="{{.ImageName}}"
666 It supports the same keys as podman inspect --format.
669 It is currently supported only by the journald log driver.
672 --mac-address=address
675 Container MAC address (e.g. 92:d0:c6:0a:29:33)
676 Remember that the MAC address in an Ethernet network must be unique.
679 The IPv6 link-local address will be based on the device's MAC address
680 according to RFC4862.
681 --memory, -m=limit
684 Memory limit (format: <number>[<unit>], where unit = b (bytes), k
685 (kilobytes), m (megabytes), or g (gigabytes))
686 Allows you to constrain the memory available to a container. If the
689 host supports swap memory, then the -m memory setting can be larger
690 than physical RAM. If a limit of 0 is specified (not using -m), the
691 container's memory is not limited. The actual limit may be rounded up
692 to a multiple of the operating system's page size (the value would be
693 very large, that's millions of trillions).
694 --memory-reservation=limit
697 Memory soft limit (format: <number>[<unit>], where unit = b (bytes), k
698 (kilobytes), m (megabytes), or g (gigabytes))
699 After setting memory reservation, when the system detects memory con‐
702 tention or low memory, containers are forced to restrict their consump‐
703 tion to their reservation. So you should always set the value below
704 --memory, otherwise the hard limit will take precedence. By default,
705 memory reservation will be the same as memory limit.
706 --memory-swap=limit
709 A limit value equal to memory plus swap. Must be used with the -m
710 (--memory) flag. The swap LIMIT should always be larger than -m (--mem‐
711 ory) value. By default, the swap LIMIT will be set to double the value
712 of --memory.
713 The format of LIMIT is <number>[<unit>]. Unit can be b (bytes), k
716 (kilobytes), m (megabytes), or g (gigabytes). If you don't specify a
717 unit, b is used. Set LIMIT to -1 to enable unlimited swap.
718 --memory-swappiness=number
721 Tune a container's memory swappiness behavior. Accepts an integer be‐
722 tween 0 and 100.
723 --mount=type=TYPE,TYPE-SPECIFIC-OPTION[,...]
726 Attach a filesystem mount to the container
727 Current supported mount TYPEs are bind, volume, image, tmpfs and de‐
730 vpts. [1] ⟨#Footnote1⟩
731 e.g.
734 type=bind,source=/path/on/host,destination=/path/in/container
736 type=bind,src=/path/on/host,dst=/path/in/container,relabel=shared
738 type=volume,source=vol1,destination=/path/in/container,ro=true
740 type=tmpfs,tmpfs-size=512M,destination=/path/in/container
742 type=image,source=fedora,destination=/fedora-image,rw=true
744 type=devpts,destination=/dev/pts
746 Common Options:
748 · src, source: mount source spec for bind and volume. Mandatory for bind.
750 · dst, destination, target: mount destination spec.
752 Options specific to volume:
754 · ro, readonly: true or false (default).
756 Options specific to image:
758 · rw, readwrite: true or false (default).
760 Options specific to bind:
762 · ro, readonly: true or false (default).
764 · bind-propagation: shared, slave, private, unbindable, rshared, rslave, runbindable, or rprivate(default). See also mount(2).
766 . bind-nonrecursive: do not setup a recursive bind mount. By default it is recursive.
768 . relabel: shared, private.
770 Options specific to tmpfs:
772 · ro, readonly: true or false (default).
774 · tmpfs-size: Size of the tmpfs mount in bytes. Unlimited by default in Linux.
776 · tmpfs-mode: File mode of the tmpfs in octal. (e.g. 700 or 0700.) Defaults to 1777 in Linux.
778 · tmpcopyup: Enable copyup from the image directory at the same location to the tmpfs. Used by default.
780 · notmpcopyup: Disable copying files from the image to the tmpfs.
782 --name=name
786 Assign a name to the container
787 The operator can identify a container in three ways: UUID long identi‐
790 fier
791 (“f78375b1c487e03c9438c729345e54db9d20cfa2ac1fc3494b6eb60872e74778”)
792 UUID short identifier (“f78375b1c487”) Name (“jonah”)
793 podman generates a UUID for each container, and if a name is not as‐
796 signed to the container with --name then it will generate a random
797 string name. The name is useful any place you need to identify a con‐
798 tainer. This works for both background and foreground containers.
799 --network=mode, --net
802 Set the network mode for the container. Invalid if using --dns,
803 --dns-opt, or --dns-search with --network that is set to none or con‐
804 tainer:id. If used together with --pod, the container will not join the
805 pod's network namespace.
806 Valid mode values are:
809 • bridge: create a network stack on the default bridge;
812 • none: no networking;
814 • container:id: reuse another container's network stack;
816 • host: use the Podman host network stack. Note: the host mode
818 gives the container full access to local system services such
819 as D-bus and is therefore considered insecure;
820 • network-id: connect to a user-defined network, multiple net‐
822 works should be comma-separated;
823 • ns:path: path to a network namespace to join;
825 • private: create a new namespace for the container (default)
827 • slirp4netns[:OPTIONS,...]: use slirp4netns(1) to create a user
829 network stack. This is the default for rootless containers. It
830 is possible to specify these additional options:
831 • allow_host_loopback=true|false: Allow the slirp4netns to
833 reach the host loopback IP (10.0.2.2). Default is false.
834 • mtu=MTU: Specify the MTU to use for this network. (Default
836 is 65520).
837 • cidr=CIDR: Specify ip range to use for this network. (De‐
839 fault is 10.0.2.0/24).
840 • enable_ipv6=true|false: Enable IPv6. Default is false. (Re‐
842 quired for outbound_addr6).
843 • outbound_addr=INTERFACE: Specify the outbound interface
845 slirp should bind to (ipv4 traffic only).
846 • outbound_addr=IPv4: Specify the outbound ipv4 address slirp
848 should bind to.
849 • outbound_addr6=INTERFACE: Specify the outbound interface
851 slirp should bind to (ipv6 traffic only).
852 • outbound_addr6=IPv6: Specify the outbound ipv6 address slirp
854 should bind to.
855 • port_handler=rootlesskit: Use rootlesskit for port forward‐
857 ing. Default.
858 • port_handler=slirp4netns: Use the slirp4netns port forward‐
860 ing.
861 --network-alias=alias
867 Add network-scoped alias for the container
868 --no-healthcheck=true|false
871 Disable any defined healthchecks for container.
872 --no-hosts=true|false
875 Do not create /etc/hosts for the container. By default, Podman will
876 manage /etc/hosts, adding the container's own IP address and any hosts
877 from --add-host.
878 --no-hosts disables this, and the image's /etc/host will be preserved un‐
881 modified.
882 This option conflicts with --add-host.
883 --oom-kill-disable=true|false
886 Whether to disable OOM Killer for the container or not.
887 --oom-score-adj=num
890 Tune the host's OOM preferences for containers (accepts -1000 to 1000)
891 --os=OS
894 Override the OS, defaults to hosts, of the image to be pulled. For ex‐
895 ample, windows.
896 --pid=pid
899 Set the PID mode for the container Default is to create a private PID
900 namespace for the container - container:<name|id>: join another con‐
901 tainer's PID namespace - host: use the host's PID namespace for the
902 container. Note: the host mode gives the container full access to local
903 PID and is therefore considered insecure. - ns: join the specified PID
904 namespace - private: create a new namespace for the container (default)
905 --pids-limit=limit
908 Tune the container's pids limit. Set 0 to have unlimited pids for the
909 container. (default "4096" on systems that support PIDS cgroups).
910 --platform=OS/ARCH
913 Specify the platform for selecting the image. (Conflicts with --arch
914 and --os) The --platform option can be used to override the current ar‐
915 chitecture and operating system.
916 --pod=name
919 Run container in an existing pod. If you want Podman to make the pod
920 for you, preference the pod name with new:. To make a pod with more
921 granular options, use the podman pod create command before creating a
922 container.
923 --pod-id-file=path
926 Run container in an existing pod and read the pod's ID from the speci‐
927 fied file. If a container is run within a pod, and the pod has an in‐
928 fra-container, the infra-container will be started before the container
929 is.
930 --privileged=true|false
933 Give extended privileges to this container. The default is false.
934 By default, Podman containers are “unprivileged” (=false) and cannot,
937 for example, modify parts of the operating system. This is because by
938 default a container is not allowed to access any devices. A “privi‐
939 leged” container is given access to all devices.
940 When the operator executes a privileged container, Podman enables ac‐
943 cess to all devices on the host, turns off graphdriver mount options,
944 as well as turning off most of the security measures protecting the
945 host from the container.
946 Rootless containers cannot have more privileges than the account that
949 launched them.
950 --publish, -p=port
953 Publish a container's port, or range of ports, to the host
954 Format: ip:hostPort:containerPort | ip::containerPort | hostPort:con‐
957 tainerPort | containerPort Both hostPort and containerPort can be spec‐
958 ified as a range of ports. When specifying ranges for both, the number
959 of container ports in the range must match the number of host ports in
960 the range. (e.g., podman run -p 1234-1236:1222-1224 --name thisWorks
961 -t busybox but not podman run -p 1230-1236:1230-1240 --name RangeCon‐
962 tainerPortsBiggerThanRangeHostPorts -t busybox) With host IP: podman
963 run -p 127.0.0.1:$HOSTPORT:$CONTAINERPORT --name CONTAINER -t someimage
964 If host IP is set to 0.0.0.0 or not set at all, the port will be bound
965 on all IPs on the host. Host port does not have to be specified (e.g.
966 podman run -p 127.0.0.1::80). If it is not, the container port will be
967 randomly assigned a port on the host. Use podman port to see the ac‐
968 tual mapping: podman port CONTAINER $CONTAINERPORT
969 Note: if a container will be run within a pod, it is not necessary to
972 publish the port for the containers in the pod. The port must only be
973 published by the pod itself. Pod network stacks act like the network
974 stack on the host - you have a variety of containers in the pod, and
975 programs in the container, all sharing a single interface and IP ad‐
976 dress, and associated ports. If one container binds to a port, no other
977 container can use that port within the pod while it is in use. Contain‐
978 ers in the pod can also communicate over localhost by having one con‐
979 tainer bind to localhost in the pod, and another connect to that port.
980 --publish-all, -P=true|false
983 Publish all exposed ports to random ports on the host interfaces. The
984 default is false.
985 When set to true publish all exposed ports to the host interfaces. The
988 default is false. If the operator uses -P (or -p) then Podman will make
989 the exposed port accessible on the host and the ports will be available
990 to any client that can reach the host. When using -P, Podman will bind
991 any exposed port to a random port on the host within an ephemeral port
992 range defined by /proc/sys/net/ipv4/ip_local_port_range. To find the
993 mapping between the host ports and the exposed ports, use podman port.
994 --pull=missing
997 Pull image before creating ("always"|"missing"|"never") (default "miss‐
998 ing").
999 'missing': default value, attempt to pull the latest image from
1000 the registries listed in registries.conf if a local image does not ex‐
1001 ist.Raise an error if the image is not in any listed registry and is
1002 not present locally.
1003 'always': Pull the image from the first registry it is found in
1004 as listed in registries.conf. Raise an error if not found in the reg‐
1005 istries, even if the image is present locally.
1006 'never': do not pull the image from the registry, use only the
1007 local version. Raise an error if the image is not present locally.
1008 Defaults to missing.
1011 --quiet, -q
1014 Suppress output information when pulling images
1015 --read-only=true|false
1018 Mount the container's root filesystem as read only.
1019 By default a container will have its root filesystem writable allowing
1022 processes to write files anywhere. By specifying the --read-only flag
1023 the container will have its root filesystem mounted as read only pro‐
1024 hibiting any writes.
1025 --read-only-tmpfs=true|false
1028 If container is running in --read-only mode, then mount a read-write
1029 tmpfs on /run, /tmp, and /var/tmp. The default is true
1030 --replace=true|false
1033 If another container with the same name already exists, replace and re‐
1034 move it. The default is false.
1035 --requires=container
1038 Specify one or more requirements. A requirement is a dependency con‐
1039 tainer that will be started before this container. Containers can be
1040 specified by name or ID, with multiple containers being separated by
1041 commas.
1042 --restart=policy
1045 Restart policy to follow when containers exit. Restart policy will not
1046 take effect if a container is stopped via the podman kill or podman
1047 stop commands.
1048 Valid values are:
1051 • no : Do not restart containers on exit
1054 • on-failure[:max_retries] : Restart containers when they exit
1056 with a non-0 exit code, retrying indefinitely or until the op‐
1057 tional max_retries count is hit
1058 • always : Restart containers when they exit,
1060 regardless of status, retrying indefinitely
1061 • unless-stopped : Identical to always
1063 Please note that restart will not restart containers after a system re‐
1067 boot. If this functionality is required in your environment, you can
1068 invoke Podman from a systemd unit file, or create an init script for
1069 whichever init system is in use. To generate systemd unit files,
1070 please see podman generate systemd
1071 --rm=true|false
1074 Automatically remove the container when it exits. The default is false.
1075 --rootfs
1078 If specified, the first argument refers to an exploded container on the
1079 file system.
1080 This is useful to run a container without requiring any image manage‐
1083 ment, the rootfs of the container is assumed to be managed externally.
1084 --sdnotify=container|conmon|ignore
1087 Determines how to use the NOTIFY_SOCKET, as passed with systemd and
1088 Type=notify.
1089 Default is container, which means allow the OCI runtime to proxy the
1092 socket into the container to receive ready notification. Podman will
1093 set the MAINPID to conmon's pid. The conmon option sets MAINPID to
1094 conmon's pid, and sends READY when the container has started. The
1095 socket is never passed to the runtime or the container. The ignore op‐
1096 tion removes NOTIFY_SOCKET from the environment for itself and child
1097 processes, for the case where some other process above Podman uses NO‐
1098 TIFY_SOCKET and Podman should not use it.
1099 --seccomp-policy=policy
1102 Specify the policy to select the seccomp profile. If set to image, Pod‐
1103 man will look for a "io.containers.seccomp.profile" label in the con‐
1104 tainer-image config and use its value as a seccomp profile. Otherwise,
1105 Podman will follow the default policy by applying the default profile
1106 unless specified otherwise via --security-opt seccomp as described be‐
1107 low.
1108 Note that this feature is experimental and may change in the future.
1111 --secret=secret[,opt=opt ...]
1114 Give the container access to a secret. Can be specified multiple times.
1115 A secret is a blob of sensitive data which a container needs at runtime
1118 but should not be stored in the image or in source control, such as
1119 usernames and passwords, TLS certificates and keys, SSH keys or other
1120 important generic strings or binary content (up to 500 kb in size).
1121 When secrets are specified as type mount, the secrets are copied and
1124 mounted into the container when a container is created. When secrets
1125 are specified as type env, the secret will be set as an environment
1126 variable within the container. Secrets are written in the container at
1127 the time of container creation, and modifying the secret using podman
1128 secret commands after the container is created will not affect the se‐
1129 cret inside the container.
1130 Secrets and its storage are managed using the podman secret command.
1133 Secret Options
1136 • type=mount|env : How the secret will be exposed to the con‐
1139 tainer. Default mount.
1140 • target=target : Target of secret. Defauts to secret name.
1142 --security-opt=option
1146 Security Options
1147 • apparmor=unconfined : Turn off apparmor confinement for the
1150 container
1151 • apparmor=your-profile : Set the apparmor confinement profile
1153 for the container
1154 • label=user:USER : Set the label user for the container
1156 processes
1157 • label=role:ROLE : Set the label role for the container
1159 processes
1160 • label=type:TYPE : Set the label process type for the con‐
1162 tainer processes
1163 • label=level:LEVEL : Set the label level for the container
1165 processes
1166 • label=filetype:TYPE : Set the label file type for the con‐
1168 tainer files
1169 • label=disable : Turn off label separation for the con‐
1171 tainer
1172 Note: Labeling can be disabled for all containers by setting la‐
1176 bel=false in the containers.conf (/etc/containers/containers.conf or
1177 $HOME/.config/containers/containers.conf) file.
1178 • mask=/path/1:/path/2 : The paths to mask separated by a colon.
1181 A masked path cannot be accessed inside the container.
1182 • no-new-privileges : Disable container processes from gaining
1184 additional privileges
1185 • seccomp=unconfined : Turn off seccomp confinement for the con‐
1187 tainer
1188 • seccomp=profile.json : White listed syscalls seccomp Json
1190 file to be used as a seccomp filter
1191 • proc-opts=OPTIONS : Comma-separated list of options to use for
1193 the /proc mount. More details for the possible mount options
1194 are specified in the proc(5) man page.
1195 • unmask=ALL or /path/1:/path/2, or shell expanded paths
1197 (/proc/*): Paths to unmask separated by a colon. If set to
1198 ALL, it will unmask all the paths that are masked or made read
1199 only by default. The default masked paths are /proc/acpi,
1200 /proc/kcore, /proc/keys, /proc/latency_stats, /proc/sched_de‐
1201 bug, /proc/scsi, /proc/timer_list, /proc/timer_stats,
1202 /sys/firmware, and /sys/fs/selinux. The default paths that
1203 are read only are /proc/asound, /proc/bus, /proc/fs,
1204 /proc/irq, /proc/sys, /proc/sysrq-trigger, /sys/fs/cgroup.
1205 Note: Labeling can be disabled for all containers by setting la‐
1209 bel=false in the containers.conf (/etc/containers/containers.conf or
1210 $HOME/.config/containers/containers.conf) file.
1211 --shm-size=size
1214 Size of /dev/shm (format: <number>[<unit>], where unit = b (bytes), k
1215 (kilobytes), m (megabytes), or g (gigabytes)) If you omit the unit, the
1216 system uses bytes. If you omit the size entirely, the system uses 64m.
1217 When size is 0, there is no limit on the amount of memory used for IPC
1218 by the container.
1219 --stop-signal=SIGTERM
1222 Signal to stop a container. Default is SIGTERM.
1223 --stop-timeout=seconds
1226 Timeout (in seconds) to stop a container. Default is 10. Remote con‐
1227 nections use local containers.conf for defaults
1228 --subgidname=name
1231 Name for GID map from the /etc/subgid file. Using this flag will run
1232 the container with user namespace enabled. This flag conflicts with
1233 --userns and --gidmap.
1234 --subuidname=name
1237 Name for UID map from the /etc/subuid file. Using this flag will run
1238 the container with user namespace enabled. This flag conflicts with
1239 --userns and --uidmap.
1240 --sysctl=SYSCTL
1243 Configure namespaced kernel parameters at runtime
1244 IPC Namespace - current sysctls allowed:
1247 kernel.msgmax, kernel.msgmnb, kernel.msgmni, kernel.sem, kernel.shmall,
1250 kernel.shmmax, kernel.shmmni, kernel.shm_rmid_forced Sysctls beginning
1251 with fs.mqueue.*
1252 Note: if you use the --ipc=host option these sysctls will not be al‐
1255 lowed.
1256 Network Namespace - current sysctls allowed:
1259 Sysctls beginning with net.*
1260 Note: if you use the --network=host option these sysctls will not be
1263 allowed.
1264 --systemd=true|false|always
1267 Run container in systemd mode. The default is true.
1268 The value always enforces the systemd mode is enforced without looking
1271 at the executable name. Otherwise, if set to true and the command you
1272 are running inside the container is systemd, /usr/sbin/init, /sbin/init
1273 or /usr/local/sbin/init.
1274 If the command you are running inside of the container is systemd, Pod‐
1277 man will setup tmpfs mount points in the following directories:
1278 /run, /run/lock, /tmp, /sys/fs/cgroup/systemd, /var/lib/journal
1281 It will also set the default stop signal to SIGRTMIN+3.
1284 This allow systemd to run in a confined container without any modifica‐
1287 tions.
1288 Note: On SELinux systems, systemd attempts to write to the cgroup file
1291 system. Containers writing to the cgroup file system are denied by de‐
1292 fault. The container_manage_cgroup boolean must be enabled for this to
1293 be allowed on an SELinux separated system.
1294 setsebool -P container_manage_cgroup true
1297 --timeout=seconds
1300 Maximum time a container is allowed to run before conmon sends it the
1301 kill signal. By default containers will run until they exit or are
1302 stopped by podman stop.
1303 --tmpfs=fs
1306 Create a tmpfs mount
1307 Mount a temporary filesystem (tmpfs) mount into a container, for exam‐
1310 ple:
1311 $ podman create -d --tmpfs /tmp:rw,size=787448k,mode=1777 my_image
1314 This command mounts a tmpfs at /tmp within the container. The supported
1317 mount options are the same as the Linux default mount flags. If you do
1318 not specify any options, the systems uses the following options:
1319 rw,noexec,nosuid,nodev.
1320 --tty, -t=true|false
1323 Allocate a pseudo-TTY. The default is false.
1324 When set to true Podman will allocate a pseudo-tty and attach to the
1327 standard input of the container. This can be used, for example, to run
1328 a throwaway interactive shell. The default is false.
1329 Note: The -t option is incompatible with a redirection of the Podman
1332 client standard input.
1333 --tz=timezone
1336 Set timezone in container. This flag takes area-based timezones, GMT
1337 time, as well as local, which sets the timezone in the container to
1338 match the host machine. See /usr/share/zoneinfo/ for valid timezones.
1339 Remote connections use local containers.conf for defaults
1340 --umask=umask
1343 Set the umask inside the container. Defaults to 0022. Remote connec‐
1344 tions use local containers.conf for defaults
1345 --uidmap=container_uid:from_uid:amount
1348 Run the container in a new user namespace using the supplied mapping.
1349 This option conflicts with the --userns and --subuidname options. This
1350 option provides a way to map host UIDs to container UIDs. It can be
1351 passed several times to map different ranges.
1352 The _fromuid value is based upon the user running the command, either
1355 rootfull or rootless users. * rootfull user: con‐
1356 tainer_uid:host_uid:amount * rootless user: container_uid:intermedi‐
1357 ate_uid:amount
1358 When podman create is called by a privileged user, the option --uidmap
1361 works as a direct mapping between host UIDs and container UIDs.
1362 host UID -> container UID
1365 The amount specifies the number of consecutive UIDs that will be
1368 mapped. If for example amount is 4 the mapping would look like:
1369 | host UID | container UID | | - | -
1372 | | _fromuid | _containeruid | | _fromuid + 1 | _containeruid +
1373 1 | | _fromuid + 2 | _containeruid + 2 | | _fromuid + 3 | _containeruid
1374 + 3 |
1375 When podman create is called by an unprivileged user (i.e. running
1378 rootless), the value _fromuid is interpreted as an "intermediate UID".
1379 In the rootless case, host UIDs are not mapped directly to container
1380 UIDs. Instead the mapping happens over two mapping steps:
1381 host UID -> intermediate UID -> container UID
1384 The --uidmap option only influences the second mapping step.
1387 The first mapping step is derived by Podman from the contents of the
1390 file /etc/subuid and the UID of the user calling Podman.
1391 First mapping step:
1394 | host UID | intermediate UID |
1397 | - | - |
1398 | UID for the user starting Podman | 0 |
1399 | 1st subordinate UID for the user starting Podman | 1 |
1400 | 2nd subordinate UID for the user starting Podman | 2 |
1401 | 3rd subordinate UID for the user starting Podman | 3 |
1402 | nth subordinate UID for the user starting Podman | n |
1403 To be able to use intermediate UIDs greater than zero, the user needs
1406 to have subordinate UIDs configured in /etc/subuid. See subuid(5).
1407 The second mapping step is configured with --uidmap.
1410 If for example amount is 5 the second mapping step would look like:
1413 | intermediate UID | container UID | | - |
1416 - | | _fromuid | _containeruid | |
1417 _fromuid + 1 | _containeruid + 1 | | _fromuid + 2 | _con‐
1418 taineruid + 2 | | _fromuid + 3 | _containeruid + 3 | | _fromuid +
1419 4 | _containeruid + 4 |
1420 Even if a user does not have any subordinate UIDs in /etc/subuid,
1423 --uidmap could still be used to map the normal UID of the user to a
1424 container UID by running podman create --uidmap $container_uid:0:1
1425 --user $container_uid ....
1426 --ulimit=option
1429 Ulimit options
1430 You can pass host to copy the current configuration from the host.
1433 --user, -u=user
1436 Sets the username or UID used and optionally the groupname or GID for
1437 the specified command.
1438 The following examples are all valid: --user [user | user:group | uid |
1441 uid:gid | user:gid | uid:group ]
1442 Without this argument the command will be run as root in the container.
1445 --userns=mode
1448 Set the user namespace mode for the container. It defaults to the POD‐
1449 MAN_USERNS environment variable. An empty value ("") means user name‐
1450 spaces are disabled unless an explicit mapping is set with the --uidmap
1451 and --gidmap options.
1452 Valid mode values are:
1455 • auto[:OPTIONS,...]: automatically create a namespace. It is
1458 possible to specify these options to auto:
1459 • gidmapping=_HOST_GID:CONTAINERGID:SIZE: to force a GID map‐
1461 ping to be present in the user namespace.
1462 • size=SIZE: to specify an explicit size for the automatic
1464 user namespace. e.g. --userns=auto:size=8192. If size is not
1465 specified, auto will estimate a size for the user namespace.
1466 • uidmapping=_HOST_UID:CONTAINERUID:SIZE: to force a UID map‐
1468 ping to be present in the user namespace.
1469 • container:id: join the user namespace of the specified con‐
1473 tainer.
1474 • host: run in the user namespace of the caller. The processes
1476 running in the container will have the same privileges on the
1477 host as any other process launched by the calling user (de‐
1478 fault).
1479 • keep-id: creates a user namespace where the current rootless
1481 user's UID:GID are mapped to the same values in the container.
1482 This option is ignored for containers created by the root
1483 user.
1484 • ns:namespace: run the container in the given existing user
1486 namespace.
1487 • private: create a new namespace for the container.
1489 This option is incompatible with --gidmap, --uidmap, --subuidname and
1493 --subgidname.
1494 --uts=mode
1497 Set the UTS namespace mode for the container. The following values are
1498 supported:
1499 • host: use the host's UTS namespace inside the container.
1502 • private: create a new namespace for the container (default).
1504 • ns:[path]: run the container in the given existing UTS name‐
1506 space.
1507 • container:[container]: join the UTS namespace of the specified
1509 container.
1510 --variant=VARIANT
1514 Use VARIANT instead of the default architecture variant of the con‐
1515 tainer image. Some images can use multiple variants of the arm archi‐
1516 tectures, such as arm/v5 and arm/v7.
1517 --volume, -v[=[[SOURCE-VOLUME|HOST-DIR:]CONTAINER-DIR[:OPTIONS]]]
1520 Create a bind mount. If you specify, -v /HOST-DIR:/CONTAINER-DIR, Pod‐
1521 man bind mounts /HOST-DIR in the host to /CONTAINER-DIR in the Podman
1522 container. Similarly, -v SOURCE-VOLUME:/CONTAINER-DIR will mount the
1523 volume in the host to the container. If no such named volume exists,
1524 Podman will create one. The OPTIONS are a comma-separated list and can
1525 be: [1] ⟨#Footnote1⟩ (Note when using the remote client, the volumes
1526 will be mounted from the remote server, not necessarily the client ma‐
1527 chine.)
1528 The options is a comma-separated list and can be:
1531 • rw|ro
1534 • z|Z
1536 • [r]shared|[r]slave|[r]private[r]unbindable
1538 • [r]bind
1540 • [no]exec
1542 • [no]dev
1544 • [no]suid
1546 • [O]
1548 • [U]
1550 The CONTAINER-DIR must be an absolute path such as /src/docs. The vol‐
1554 ume will be mounted into the container at this directory.
1555 Volumes may specify a source as well, as either a directory on the host
1558 or the name of a named volume. If no source is given, the volume will
1559 be created as an anonymously named volume with a randomly generated
1560 name, and will be removed when the container is removed via the --rm
1561 flag or podman rm --volumes.
1562 If a volume source is specified, it must be a path on the host or the
1565 name of a named volume. Host paths are allowed to be absolute or rela‐
1566 tive; relative paths are resolved relative to the directory Podman is
1567 run in. If the source does not exist, Podman will return an error.
1568 Users must pre-create the source files or directories.
1569 Any source that does not begin with a . or / will be treated as the
1572 name of a named volume. If a volume with that name does not exist, it
1573 will be created. Volumes created with names are not anonymous, and
1574 they are not removed by the --rm option and the podman rm --volumes
1575 command.
1576 You can specify multiple -v options to mount one or more volumes into
1579 a container.
1580 Write Protected Volume Mounts
1583 You can add :ro or :rw suffix to a volume to mount it read-only or
1586 read-write mode, respectively. By default, the volumes are mounted
1587 read-write. See examples.
1588 Chowning Volume Mounts
1591 By default, Podman does not change the owner and group of source volume
1594 directories mounted into containers. If a container is created in a new
1595 user namespace, the UID and GID in the container may correspond to an‐
1596 other UID and GID on the host.
1597 The :U suffix tells Podman to use the correct host UID and GID based on
1600 the UID and GID within the container, to change recursively the owner
1601 and group of the source volume.
1602 Warning use with caution since this will modify the host filesystem.
1605 Labeling Volume Mounts
1608 Labeling systems like SELinux require that proper labels are placed on
1611 volume content mounted into a container. Without a label, the security
1612 system might prevent the processes running inside the container from
1613 using the content. By default, Podman does not change the labels set by
1614 the OS.
1615 To change a label in the container context, you can add either of two
1618 suffixes :z or :Z to the volume mount. These suffixes tell Podman to
1619 relabel file objects on the shared volumes. The z option tells Podman
1620 that two containers share the volume content. As a result, Podman la‐
1621 bels the content with a shared content label. Shared volume labels al‐
1622 low all containers to read/write content. The Z option tells Podman to
1623 label the content with a private unshared label. Only the current con‐
1624 tainer can use a private volume.
1625 Overlay Volume Mounts
1628 The :O flag tells Podman to mount the directory from the host as a tem‐
1631 porary storage using the overlay file system. The container processes
1632 can modify content within the mountpoint which is stored in the con‐
1633 tainer storage in a separate directory. In overlay terms, the source
1634 directory will be the lower, and the container storage directory will
1635 be the upper. Modifications to the mount point are destroyed when the
1636 container finishes executing, similar to a tmpfs mount point being un‐
1637 mounted.
1638 Subsequent executions of the container will see the original source di‐
1641 rectory content, any changes from previous container executions no
1642 longer exist.
1643 One use case of the overlay mount is sharing the package cache from the
1646 host into the container to allow speeding up builds.
1647 Note:
1650 - The `O` flag conflicts with other options listed above.
1653 Content mounted into the container is labeled with the private label.
1657 On SELinux systems, labels in the source directory must be read‐
1658 able by the container label. Usually containers can read/execute con‐
1659 tainer_share_t and can read/write container_file_t. If you cannot
1660 change the labels on a source volume, SELinux container separation must
1661 be disabled for the container to work.
1662 - The source directory mounted into the container with an overlay
1663 mount should not be modified, it can cause unexpected failures. It is
1664 recommended that you do not modify the directory until the container
1665 finishes running.
1666 Mounts propagation
1669 By default bind mounted volumes are private. That means any mounts done
1672 inside container will not be visible on host and vice versa. One can
1673 change this behavior by specifying a volume mount propagation property.
1674 Making a volume shared mounts done under that volume inside container
1675 will be visible on host and vice versa. Making a volume slave enables
1676 only one way mount propagation and that is mounts done on host under
1677 that volume will be visible inside container but not the other way
1678 around. [1] ⟨#Footnote1⟩
1679 To control mount propagation property of a volume one can use the
1682 [r]shared, [r]slave, [r]private or the [r]unbindable propagation flag.
1683 Propagation property can be specified only for bind mounted volumes and
1684 not for internal volumes or named volumes. For mount propagation to
1685 work the source mount point (the mount point where source dir is
1686 mounted on) has to have the right propagation properties. For shared
1687 volumes, the source mount point has to be shared. And for slave vol‐
1688 umes, the source mount point has to be either shared or slave. [1]
1689 ⟨#Footnote1⟩
1690 If you want to recursively mount a volume and all of its submounts into
1693 a container, then you can use the rbind option. By default the bind op‐
1694 tion is used, and submounts of the source directory will not be mounted
1695 into the container.
1696 Mounting the volume with the nosuid options means that SUID applica‐
1699 tions on the volume will not be able to change their privilege. By de‐
1700 fault volumes are mounted with nosuid.
1701 Mounting the volume with the noexec option means that no executables on
1704 the volume will be able to executed within the container.
1705 Mounting the volume with the nodev option means that no devices on the
1708 volume will be able to be used by processes within the container. By
1709 default volumes are mounted with nodev.
1710 If the <source-dir> is a mount point, then "dev", "suid", and "exec"
1713 options are ignored by the kernel.
1714 Use df <source-dir> to figure out the source mount and then use findmnt
1717 -o TARGET,PROPAGATION <source-mount-dir> to figure out propagation
1718 properties of source mount. If findmnt utility is not available, then
1719 one can look at mount entry for source mount point in
1720 /proc/self/mountinfo. Look at optional fields and see if any propaga‐
1721 tion properties are specified. shared:X means mount is shared, mas‐
1722 ter:X means mount is slave and if nothing is there that means mount is
1723 private. [1] ⟨#Footnote1⟩
1724 To change propagation properties of a mount point use mount command.
1727 For example, if one wants to bind mount source directory /foo one can
1728 do mount --bind /foo /foo and mount --make-private --make-shared /foo.
1729 This will convert /foo into a shared mount point. Alternatively one can
1730 directly change propagation properties of source mount. Say / is source
1731 mount for /foo, then use mount --make-shared / to convert / into a
1732 shared mount.
1733 Note: if the user only has access rights via a group, accessing the
1736 volume from inside a rootless container will fail. Use the --group-add
1737 keep-groups flag to pass the user's supplementary group access into the
1738 container.
1739 --volumes-from[=CONTAINER[:OPTIONS]]
1742 Mount volumes from the specified container(s). Used to share volumes
1743 between containers. The options is a comma-separated list with the fol‐
1744 lowing available elements:
1745 • rw|ro
1748 • z
1750 Mounts already mounted volumes from a source container onto another
1754 container. You must supply the source's container-id or container-name.
1755 To share a volume, use the --volumes-from option when running the tar‐
1756 get container. You can share volumes even if the source container is
1757 not running.
1758 By default, Podman mounts the volumes in the same mode (read-write or
1761 read-only) as it is mounted in the source container. You can change
1762 this by adding a ro or rw option.
1763 Labeling systems like SELinux require that proper labels are placed on
1766 volume content mounted into a container. Without a label, the security
1767 system might prevent the processes running inside the container from
1768 using the content. By default, Podman does not change the labels set by
1769 the OS.
1770 To change a label in the container context, you can add z to the volume
1773 mount. This suffix tells Podman to relabel file objects on the shared
1774 volumes. The z option tells Podman that two containers share the volume
1775 content. As a result, Podman labels the content with a shared content
1776 label. Shared volume labels allow all containers to read/write content.
1777 If the location of the volume from the source container overlaps with
1780 data residing on a target container, then the volume hides that data on
1781 the target.
1782 --workdir, -w=dir
1785 Working directory inside the container
1786 The default working directory for running binaries within a container
1789 is the root directory (/). The image developer can set a different de‐
1790 fault with the WORKDIR instruction. The operator can override the work‐
1791 ing directory by using the -w option.
1792 --pidfile=path
1795 When the pidfile location is specified, the container process' PID will
1796 be written to the pidfile. (This option is not available with the re‐
1797 mote Podman client) If the pidfile option is not specified, the con‐
1798 tainer process' PID will be written to /run/containers/storage/${stor‐
1799 age-driver}-containers/$CID/userdata/pidfile.
1800 After the container is started, the location for the pidfile can be
1803 discovered with the following podman inspect command:
1804 $ podman inspect --format '{{ .PidFile }}' $CID
1807 /run/containers/storage/${storage-driver}-containers/$CID/userdata/pidfile
1808
1812 Create a container using a local image
1813 $ podman create alpine ls
1814 Create a container using a local image and annotate it
1818 $ podman create --annotation HELLO=WORLD alpine ls
1819 Create a container using a local image, allocating a pseudo-TTY, keeping
1823 stdin open and name it myctr
1824 podman create -t -i --name myctr alpine ls
1825 Set UID/GID mapping in a new user namespace
1829 Running a container in a new user namespace requires a mapping of the
1830 uids and gids from the host.
1831 $ podman create --uidmap 0:30000:7000 --gidmap 0:30000:7000 fedora echo hello
1834 Configure timezone in a container
1838 $ podman create --tz=local alpine date
1839 $ podman create --tz=Asia/Shanghai alpine date
1840 $ podman create --tz=US/Eastern alpine date
1841 Adding dependency containers
1845 Podman will make sure the first container, container1, is running be‐
1846 fore the second container (container2) is started.
1847 $ podman create --name container1 -t -i fedora bash
1850 $ podman create --name container2 --requires container1 -t -i fedora bash
1851 $ podman start --attach container2
1852 Multiple containers can be required.
1856 $ podman create --name container1 -t -i fedora bash
1859 $ podman create --name container2 -t -i fedora bash
1860 $ podman create --name container3 --requires container1,container2 -t -i fedora bash
1861 $ podman start --attach container3
1862 Configure keep supplemental groups for access to volume
1866 $ podman create -v /var/lib/design:/var/lib/design --group-add keep-groups ubi8
1867 Rootless Containers
1871 Podman runs as a non root user on most systems. This feature requires
1872 that a new enough version of shadow-utils be installed. The
1873 shadow-utils package must include the newuidmap and newgidmap executa‐
1874 bles.
1875 Note: RHEL7 and Centos 7 will not have this feature until RHEL7.7 is
1878 released.
1879 In order for users to run rootless, there must be an entry for their
1882 username in /etc/subuid and /etc/subgid which lists the UIDs for their
1883 user namespace.
1884 Rootless Podman works better if the fuse-overlayfs and slirp4netns
1887 packages are installed. The fuse-overlay package provides a userspace
1888 overlay storage driver, otherwise users need to use the vfs storage
1889 driver, which is diskspace expensive and does not perform well.
1890 slirp4netns is required for VPN, without it containers need to be run
1891 with the --network=host flag.
1892
1895 Environment variables within containers can be set using multiple dif‐
1896 ferent options: This section describes the precedence.
1897 Precedence order (later entries override earlier entries):
1900 • --env-host : Host environment of the process executing Podman
1903 is added.
1904 • --http-proxy: By default, several environment variables will
1906 be passed in from the host, such as http_proxy and no_proxy.
1907 See --http-proxy for details.
1908 • Container image : Any environment variables specified in the
1910 container image.
1911 • --env-file : Any environment variables specified via
1913 env-files. If multiple files specified, then they override
1914 each other in order of entry.
1915 • --env : Any environment variables specified will override pre‐
1917 vious settings.
1918 Create containers and set the environment ending with a * and a *****
1922 $ export ENV1=a
1925 $ podman create --name ctr --env ENV* alpine printenv ENV1
1926 $ podman start --attach ctr
1927 a
1928 $ podman create --name ctr --env ENV*****=b alpine printenv ENV*****
1930 $ podman start --attach ctr
1931 b
1932
1936 When Podman starts a container it actually executes the conmon program,
1937 which then executes the OCI Runtime. Conmon is the container monitor.
1938 It is a small program whose job is to watch the primary process of the
1939 container, and if the container dies, save the exit code. It also
1940 holds open the tty of the container, so that it can be attached to
1941 later. This is what allows Podman to run in detached mode (back‐
1942 grounded), so Podman can exit but conmon continues to run. Each con‐
1943 tainer has their own instance of conmon. Conmon waits for the container
1944 to exit, gathers and saves the exit code, and then launches a Podman
1945 process to complete the container cleanup, by shutting down the network
1946 and storage. For more information on conmon, please reference the
1947 conmon(8) man page.
1948
1951 /etc/subuid /etc/subgid
1952 NOTE: Use the environment variable TMPDIR to change the temporary stor‐
1955 age location of downloaded container images. Podman defaults to use
1956 /var/tmp.
1957
1960 podman(1), podman-secret(1), podman-save(1), podman-ps(1), podman-at‐
1961 tach(1), podman-pod-create(1), podman-port(1), *podman-start(1), pod‐
1962 man-kill(1), podman-stop(1), podman-generate-systemd(1) podman-rm(1),
1963 subgid(5), subuid(5), containers.conf(5), systemd.unit(5), setse‐
1964 bool(8), slirp4netns(1), fuse-overlayfs(1), proc(5), conmon(8).
1965
1968 October 2017, converted from Docker documentation to Podman by Dan
1969 Walsh for Podman <dwalsh@redhat.com>
1970 November 2014, updated by Sven Dowideit <SvenDowideit@home.org.au>
1973 September 2014, updated by Sven Dowideit <SvenDowideit@home.org.au>
1976 August 2014, updated by Sven Dowideit <SvenDowideit@home.org.au>
1979
1982 1: The Podman project is committed to inclusivity, a core value of open
1983 source. The master and slave mount propagation terminology used here is
1984 problematic and divisive, and should be changed. However, these terms
1985 are currently used within the Linux kernel and must be used as-is at
1986 this time. When the kernel maintainers rectify this usage, Podman will
1987 follow suit immediately.
1988 podman-create(1)()