Docker run reference
Docker runs processes in isolated containers. A container is a process which runs on a host. The host may be local or remote. When an operator executes , the container process that runs is isolated in that it has its own file system, its own networking, and its own isolated process tree separate from the host.
This page details how to use the command to define the container’s resources at runtime.
The basic command takes this form:
The command must specify an IMAGE to derive the container from. An image developer can define image defaults related to:
detached or foreground running container identification network settings runtime constraints on CPU and memory With the an operator can add to or override the image defaults set by a developer. And, additionally, operators can override nearly all the defaults set by the Docker runtime itself. The operator’s ability to override image and Docker runtime defaults is why run has more options than any other command.
To learn how to interpret the types of , see Option types.
Note
Depending on your Docker system configuration, you may be required to preface the command with . To avoid having to use with the command, your system administrator can create a Unix group called and add users to it. For more information about this configuration, refer to the Docker installation documentation for your operating system.
Only the operator (the person executing ) can set the following options.
Detached vs foreground Detached (-d) Foreground Container identification Name (--name) PID equivalent IPC settings (--ipc) Network settings Restart policies (--restart) Clean up (--rm) Runtime constraints on resources Runtime privilege and Linux capabilities When starting a Docker container, you must first decide if you want to run the container in the background in a “detached” mode or in the default foreground mode:
To start a container in detached mode, you use or just option. By design, containers started in detached mode exit when the root process used to run the container exits, unless you also specify the option. If you use with , the container is removed when it exits or when the daemon exits, whichever happens first.
Do not pass a command to a detached container. For example, this command attempts to start the service.
This succeeds in starting the service inside the container. However, it fails the detached container paradigm in that, the root process () returns and the detached container stops as designed. As a result, the service is started but could not be used. Instead, to start a process such as the web server do the following:
To do input/output with a detached container use network connections or shared volumes. These are required because the container is no longer listening to the command line where was run.
To reattach to a detached container, use attach command.
In foreground mode (the default when is not specified), can start the process in the container and attach the console to the process’s standard input, output, and standard error. It can even pretend to be a TTY (this is what most command line executables expect) and pass along signals. All of that is configurable:
If you do not specify then Docker will attach to both stdout and stderr . You can specify to which of the three standard streams (, , ) you’d like to connect instead, as in:
For interactive processes (like a shell), you must use together in order to allocate a tty for the container process. is often written as you’ll see in later examples. Specifying is forbidden when the client is receiving its standard input from a pipe, as in:
Note
A process running as PID 1 inside a container is treated specially by Linux: it ignores any signal with the default action. As a result, the process will not terminate on or unless it is coded to do so.
The operator can identify a container in three ways:
Identifier type Example value UUID long identifier “f78375b1c487e03c9438c729345e54db9d20cfa2ac1fc3494b6eb60872e74778” UUID short identifier “f78375b1c487” Name “evil_ptolemy” The UUID identifiers come from the Docker daemon. If you do not assign a container name with the option, then the daemon generates a random string name for you. Defining a can be a handy way to add meaning to a container. If you specify a , you can use it when referencing the container within a Docker network. This works for both background and foreground Docker containers.
Note
Containers on the default bridge network must be linked to communicate by name.
Finally, to help with automation, you can have Docker write the container ID out to a file of your choosing. This is similar to how some programs might write out their process ID to a file (you’ve seen them as PID files):
While not strictly a means of identifying a container, you can specify a version of an image you’d like to run the container with by adding to the command. For example, .
Images using the v2 or later image format have a content-addressable identifier called a digest. As long as the input used to generate the image is unchanged, the digest value is predictable and referenceable.
The following example runs a container from the image with the digest:
By default, all containers have the PID namespace enabled.
PID namespace provides separation of processes. The PID Namespace removes the view of the system processes, and allows process ids to be reused including pid 1.
In certain cases you want your container to share the host’s process namespace, basically allowing processes within the container to see all of the processes on the system. For example, you could build a container with debugging tools like or , but want to use these tools when debugging processes within the container.
Create this Dockerfile:
Build the Dockerfile and tag the image as :
Use the following command to run inside a container:
Joining another container’s pid namespace can be used for debugging that container.
Start a container running a redis server:
Debug the redis container by running another container that has strace in it:
The UTS namespace is for setting the hostname and the domain that is visible to running processes in that namespace. By default, all containers, including those with , have their own UTS namespace. The setting will result in the container using the same UTS namespace as the host. Note that and are invalid in UTS mode.
You may wish to share the UTS namespace with the host if you would like the hostname of the container to change as the hostname of the host changes. A more advanced use case would be changing the host’s hostname from a container.
The following values are accepted:
Value Description ”” Use daemon’s default. “none” Own private IPC namespace, with /dev/shm not mounted. “private” Own private IPC namespace. “shareable” Own private IPC namespace, with a possibility to share it with other containers. “container: " Join another (“shareable”) container’s IPC namespace. “host” Use the host system’s IPC namespace. If not specified, daemon default is used, which can either be or , depending on the daemon version and configuration.
IPC (POSIX/SysV IPC) namespace provides separation of named shared memory segments, semaphores and message queues.
Shared memory segments are used to accelerate inter-process communication at memory speed, rather than through pipes or through the network stack. Shared memory is commonly used by databases and custom-built (typically C/OpenMPI, C++/using boost libraries) high performance applications for scientific computing and financial services industries. If these types of applications are broken into multiple containers, you might need to share the IPC mechanisms of the containers, using mode for the main (i.e. “donor”) container, and for other containers.
By default, all containers have networking enabled and they can make any outgoing connections. The operator can completely disable networking with which disables all incoming and outgoing networking. In cases like this, you would perform I/O through files or and only.
Publishing ports and linking to other containers only works with the default (bridge). The linking feature is a legacy feature. You should always prefer using Docker network drivers over linking.
Your container will use the same DNS servers as the host by default, but you can override this with .
By default, the MAC address is generated using the IP address allocated to the container. You can set the container’s MAC address explicitly by providing a MAC address via the parameter (format:).Be aware that Docker does not check if manually specified MAC addresses are unique.
Supported networks :
Description No networking in the container. Connect the container to the bridge via veth interfaces. Use the host's network stack inside the container. Use the network stack of another container, specified via its name or id. Connects the container to a user created network (using command) Network: none With the network is a container will not have access to any external routes. The container will still have a interface enabled in the container but it does not have any routes to external traffic.
Network: bridge With the network set to a container will use docker’s default networking setup. A bridge is setup on the host, commonly named , and a pair of interfaces will be created for the container. One side of the pair will remain on the host attached to the bridge while the other side of the pair will be placed inside the container’s namespaces in addition to the interface. An IP address will be allocated for containers on the bridge’s network and traffic will be routed though this bridge to the container.
Containers can communicate via their IP addresses by default. To communicate by name, they must be linked.
Network: host With the network set to a container will share the host’s network stack and all interfaces from the host will be available to the container. The container’s hostname will match the hostname on the host system. Note that is invalid in netmode. Even in network mode a container has its own UTS namespace by default. As such and are allowed in network mode and will only change the hostname and domain name inside the container. Similar to , the , , , and options can be used in network mode. These options update or inside the container. No change are made to and on the host.
Compared to the default mode, the mode gives significantly better networking performance since it uses the host’s native networking stack whereas the bridge has to go through one level of virtualization through the docker daemon. It is recommended to run containers in this mode when their networking performance is critical, for example, a production Load Balancer or a High Performance Web Server.
Note
gives the container full access to local system services such as D-bus and is therefore considered insecure.
Network: container With the network set to a container will share the network stack of another container. The other container’s name must be provided in the format of . Note that and are invalid in netmode, and are also invalid in netmode.
Example running a Redis container with Redis binding to then running the command and connecting to the Redis server over the interface.
User-defined network You can create a network using a Docker network driver or an external network driver plugin. You can connect multiple containers to the same network. Once connected to a user-defined network, the containers can communicate easily using only another container’s IP address or name.
For networks or custom plugins that support multi-host connectivity, containers connected to the same multi-host network but launched from different Engines can also communicate in this way.
The following example creates a network using the built-in network driver and running a container in the created network
Your container will have lines in which define the hostname of the container itself as well as and a few other common things. The flag can be used to add additional lines to .
If a container is connected to the default bridge network and with other containers, then the container’s file is updated with the linked container’s name.
Note
Since Docker may live update the container’s file, there may be situations when processes inside the container can end up reading an empty or incomplete file. In most cases, retrying the read again should fix the problem.
Using the flag on Docker run you can specify a restart policy for how a container should or should not be restarted on exit.
When a restart policy is active on a container, it will be shown as either or in . It can also be useful to use to see the restart policy in effect.
Docker supports the following restart policies:
Policy Result no Do not automatically restart the container when it exits. This is the default. on-failure[:max-retries] Restart only if the container exits with a non-zero exit status. Optionally, limit the number of restart retries the Docker daemon attempts. always Always restart the container regardless of the exit status. When you specify always, the Docker daemon will try to restart the container indefinitely. The container will also always start on daemon startup, regardless of the current state of the container. unless-stopped Always restart the container regardless of the exit status, including on daemon startup, except if the container was put into a stopped state before the Docker daemon was stopped. An increasing delay (double the previous delay, starting at 100 milliseconds) is added before each restart to prevent flooding the server. This means the daemon will wait for 100 ms, then 200 ms, 400, 800, 1600, and so on until either the limit, the maximum delay of 1 minute is hit, or when you or the container.
If a container is successfully restarted (the container is started and runs for at least 10 seconds), the delay is reset to its default value of 100 ms.
You can specify the maximum amount of times Docker will try to restart the container when using the on-failure policy. The default is that Docker will try forever to restart the container. The number of (attempted) restarts for a container can be obtained via . For example, to get the number of restarts for container “my-container”;
Or, to get the last time the container was (re)started;
Combining (restart policy) with the (clean up) flag results in an error. On container restart, attached clients are disconnected. See the examples on using the (clean up) flag later in this page.
This will run the container with a restart policy of always so that if the container exits, Docker will restart it.
This will run the container with a restart policy of on-failure and a maximum restart count of 10. If the container exits with a non-zero exit status more than 10 times in a row Docker will abort trying to restart the container. Providing a maximum restart limit is only valid for the on-failure policy.
The exit code from gives information about why the container failed to run or why it exited. When exits with a non-zero code, the exit codes follow the standard, see below:
125 if the error is with Docker daemon itself
126 if the contained command cannot be invoked
127 if the contained command cannot be found
Exit code of contained command otherwise
By default a container’s file system persists even after the container exits. This makes debugging a lot easier (since you can inspect the final state) and you retain all your data by default. But if you are running short-term foreground processes, these container file systems can really pile up. If instead you’d like Docker to automatically clean up the container and remove the file system when the container exits, you can add the flag:
Note
If you set the flag, Docker also removes the anonymous volumes associated with the container when the container is removed. This is similar to running . Only volumes that are specified without a name are removed. For example, when running:
the volume for will be removed, but the volume for will not. Volumes inherited via will be removed with the same logic: if the original volume was specified with a name it will not be removed.
Option Description Set the label user for the container Set the label role for the container Set the label type for the container Set the label level for the container Turn off label confinement for the container Set the apparmor profile to be applied to the container Disable container processes from gaining new privileges Turn off seccomp confinement for the container White-listed syscalls seccomp Json file to be used as a seccomp filter You can override the default labeling scheme for each container by specifying the flag. Specifying the level in the following command allows you to share the same content between containers.
Note
Automatic translation of MLS labels is not currently supported.
To disable the security labeling for this container versus running with the flag, use the following command:
If you want a tighter security policy on the processes within a container, you can specify an alternate type for the container. You could run a container that is only allowed to listen on Apache ports by executing the following command:
Note
You would have to write policy defining a type.
If you want to prevent your container processes from gaining additional privileges, you can execute the following command:
This means that commands that raise privileges such as or will no longer work. It also causes any seccomp filters to be applied later, after privileges have been dropped which may mean you can have a more restrictive set of filters. For more details, see the kernel documentation.
You can use the flag to indicate that an init process should be used as the PID 1 in the container. Specifying an init process ensures the usual responsibilities of an init system, such as reaping zombie processes, are performed inside the created container.
The default init process used is the first executable found in the system path of the Docker daemon process. This binary, included in the default installation, is backed by tini.
Using the flag, you can pass a specific cgroup to run a container in. This allows you to create and manage cgroups on their own. You can define custom resources for those cgroups and put containers under a common parent group.
The operator can also adjust the performance parameters of the container:
Option Description , Memory limit (format: ). Number is a positive integer. Unit can be one of , , , or . Minimum is 6M. Total memory limit (memory + swap, format: ). Number is a positive integer. Unit can be one of , , , or . Memory soft limit (format: ). Number is a positive integer. Unit can be one of , , , or . Kernel memory limit (format: ). Number is a positive integer. Unit can be one of , , , or . Minimum is 4M. , CPU shares (relative weight) Number of CPUs. Number is a fractional number. 0.000 means no limit. Limit the CPU CFS (Completely Fair Scheduler) period CPUs in which to allow execution (0-3, 0,1) Memory nodes (MEMs) in which to allow execution (0-3, 0,1). Only effective on NUMA systems. Limit the CPU CFS (Completely Fair Scheduler) quota Limit the CPU real-time period. In microseconds. Requires parent cgroups be set and cannot be higher than parent. Also check rtprio ulimits. Limit the CPU real-time runtime. In microseconds. Requires parent cgroups be set and cannot be higher than parent. Also check rtprio ulimits. Block IO weight (relative weight) accepts a weight value between 10 and 1000. Block IO weight (relative device weight, format: ) Limit read rate from a device (format: ). Number is a positive integer. Unit can be one of , , or . Limit write rate to a device (format: ). Number is a positive integer. Unit can be one of , , or . Limit read rate (IO per second) from a device (format: ). Number is a positive integer. Limit write rate (IO per second) to a device (format: ). Number is a positive integer. Whether to disable OOM Killer for the container or not. Tune container’s OOM preferences (-1000 to 1000) Tune a container’s memory swappiness behavior. Accepts an integer between 0 and 100. Size of . The format is . must be greater than . Unit is optional and can be (bytes), (kilobytes), (megabytes), or (gigabytes). If you omit the unit, the system uses bytes. If you omit the size entirely, the system uses . We have four ways to set user memory usage:
Option Result There is no memory limit for the container. The container can use as much memory as needed. (specify memory and set memory-swap as ) The container is not allowed to use more than L bytes of memory, but can use as much swap as is needed (if the host supports swap memory). (specify memory without memory-swap) The container is not allowed to use more than L bytes of memory, swap plus memory usage is double of that. (specify both memory and memory-swap) The container is not allowed to use more than L bytes of memory, swap plus memory usage is limited by S. Examples:
We set nothing about memory, this means the processes in the container can use as much memory and swap memory as they need.
We set memory limit and disabled swap memory limit, this means the processes in the container can use 300M memory and as much swap memory as they need (if the host supports swap memory).
We set memory limit only, this means the processes in the container can use 300M memory and 300M swap memory, by default, the total virtual memory size (--memory-swap) will be set as double of memory, in this case, memory + swap would be 2*300M, so processes can use 300M swap memory as well.
We set both memory and swap memory, so the processes in the container can use 300M memory and 700M swap memory.
Memory reservation is a kind of memory soft limit that allows for greater sharing of memory. Under normal circumstances, containers can use as much of the memory as needed and are constrained only by the hard limits set with the / option. When memory reservation is set, Docker detects memory contention or low memory and forces containers to restrict their consumption to a reservation limit.
Always set the memory reservation value below the hard limit, otherwise the hard limit takes precedence. A reservation of 0 is the same as setting no reservation. By default (without reservation set), memory reservation is the same as the hard memory limit.
Memory reservation is a soft-limit feature and does not guarantee the limit won’t be exceeded. Instead, the feature attempts to ensure that, when memory is heavily contended for, memory is allocated based on the reservation hints/setup.
The following example limits the memory () to 500M and sets the memory reservation to 200M.
Under this configuration, when the container consumes memory more than 200M and less than 500M, the next system memory reclaim attempts to shrink container memory below 200M.
The following example set memory reservation to 1G without a hard memory limit.
The container can use as much memory as it needs. The memory reservation setting ensures the container doesn’t consume too much memory for long time, because every memory reclaim shrinks the container’s consumption to the reservation.
By default, kernel kills processes in a container if an out-of-memory (OOM) error occurs. To change this behaviour, use the option. Only disable the OOM killer on containers where you have also set the option. If the flag is not set, this can result in the host running out of memory and require killing the host’s system processes to free memory.
The following example limits the memory to 100M and disables the OOM killer for this container:
The following example, illustrates a dangerous way to use the flag:
The container has unlimited memory which can cause the host to run out memory and require killing system processes to free memory. The parameter can be changed to select the priority of which containers will be killed when the system is out of memory, with negative scores making them less likely to be killed, and positive scores more likely.
Kernel memory is fundamentally different than user memory as kernel memory can’t be swapped out. The inability to swap makes it possible for the container to block system services by consuming too much kernel memory. Kernel memory includes：
stack pages slab pages sockets memory pressure tcp memory pressure You can setup kernel memory limit to constrain these kinds of memory. For example, every process consumes some stack pages. By limiting kernel memory, you can prevent new processes from being created when the kernel memory usage is too high.
Kernel memory is never completely independent of user memory. Instead, you limit kernel memory in the context of the user memory limit. Assume “U” is the user memory limit and “K” the kernel limit. There are three possible ways to set limits:
Option Result This is the standard memory limitation mechanism already present before using kernel memory. Kernel memory is completely ignored. Kernel memory is a subset of the user memory. This setup is useful in deployments where the total amount of memory per-cgroup is overcommitted. Overcommitting kernel memory limits is definitely not recommended, since the box can still run out of non-reclaimable memory. In this case, you can configure K so that the sum of all groups is never greater than the total memory. Then, freely set U at the expense of the system's service quality. Since kernel memory charges are also fed to the user counter and reclamation is triggered for the container for both kinds of memory. This configuration gives the admin a unified view of memory. It is also useful for people who just want to track kernel memory usage. Examples:
We set memory and kernel memory, so the processes in the container can use 500M memory in total, in this 500M memory, it can be 50M kernel memory tops.
We set kernel memory without -m, so the processes in the container can use as much memory as they want, but they can only use 50M kernel memory.
By default, a container’s kernel can swap out a percentage of anonymous pages. To set this percentage for a container, specify a value between 0 and 100. A value of 0 turns off anonymous page swapping. A value of 100 sets all anonymous pages as swappable. By default, if you are not using , memory swappiness value will be inherited from the parent.
For example, you can set:
Setting the option is helpful when you want to retain the container’s working set and to avoid swapping performance penalties.
By default, all containers get the same proportion of CPU cycles. This proportion can be modified by changing the container’s CPU share weighting relative to the weighting of all other running containers.
To modify the proportion from the default of 1024, use the or flag to set the weighting to 2 or higher. If 0 is set, the system will ignore the value and use the default of 1024.
The proportion will only apply when CPU-intensive processes are running. When tasks in one container are idle, other containers can use the left-over CPU time. The actual amount of CPU time will vary depending on the number of containers running on the system.
For example, consider three containers, one has a cpu-share of 1024 and two others have a cpu-share setting of 512. When processes in all three containers attempt to use 100% of CPU, the first container would receive 50% of the total CPU time. If you add a fourth container with a cpu-share of 1024, the first container only gets 33% of the CPU. The remaining containers receive 16.5%, 16.5% and 33% of the CPU.
On a multi-core system, the shares of CPU time are distributed over all CPU cores. Even if a container is limited to less than 100% of CPU time, it can use 100% of each individual CPU core.
For example, consider a system with more than three cores. If you start one container with running one process, and another container with running two processes, this can result in the following division of CPU shares:
The default CPU CFS (Completely Fair Scheduler) period is 100ms. We can use to set the period of CPUs to limit the container’s CPU usage. And usually should work with .
Examples:
If there is 1 CPU, this means the container can get 50% CPU worth of run-time every 50ms.
In addition to use and for setting CPU period constraints, it is possible to specify with a float number to achieve the same purpose. For example, if there is 1 CPU, then will achieve the same result as setting and (50% CPU).
The default value for is , which means there is no limit.
For more information, see the CFS documentation on bandwidth limiting.
We can set cpus in which to allow execution for containers.
Examples:
This means processes in container can be executed on cpu 1 and cpu 3.
This means processes in container can be executed on cpu 0, cpu 1 and cpu 2.
We can set mems in which to allow execution for containers. Only effective on NUMA systems.
Examples:
This example restricts the processes in the container to only use memory from memory nodes 1 and 3.
This example restricts the processes in the container to only use memory from memory nodes 0, 1 and 2.
The flag limits the container’s CPU usage. The default 0 value allows the container to take 100% of a CPU resource (1 CPU). The CFS (Completely Fair Scheduler) handles resource allocation for executing processes and is default Linux Scheduler used by the kernel. Set this value to 50000 to limit the container to 50% of a CPU resource. For multiple CPUs, adjust the as necessary. For more information, see the CFS documentation on bandwidth limiting.
By default, all containers get the same proportion of block IO bandwidth (blkio). This proportion is 500. To modify this proportion, change the container’s blkio weight relative to the weighting of all other running containers using the flag.
Note:
The blkio weight setting is only available for direct IO. Buffered IO is not currently supported.
The flag can set the weighting to a value between 10 to 1000. For example, the commands below create two containers with different blkio weight:
If you do block IO in the two containers at the same time, by, for example:
You’ll find that the proportion of time is the same as the proportion of blkio weights of the two containers.
The flag sets a specific device weight. The is a string containing a colon-separated device name and weight. For example, to set device weight to :
If you specify both the and , Docker uses the as the default weight and uses to override this default with a new value on a specific device. The following example uses a default weight of and overrides this default on setting that weight to :
The flag limits the read rate (bytes per second) from a device. For example, this command creates a container and limits the read rate to per second from :
The flag limits the write rate (bytes per second) to a device. For example, this command creates a container and limits the write rate to per second for :
Both flags take limits in the format. Both read and write rates must be a positive integer. You can specify the rate in (kilobytes), (megabytes), or (gigabytes).
The flag limits read rate (IO per second) from a device. For example, this command creates a container and limits the read rate to IO per second from :
The flag limits write rate (IO per second) to a device. For example, this command creates a container and limits the write rate to IO per second to :
Both flags take limits in the format. Both read and write rates must be a positive integer.
By default, the docker container process runs with the supplementary groups looked up for the specified user. If one wants to add more to that list of groups, then one can use this flag:
Option Description Add Linux capabilities Drop Linux capabilities Give extended privileges to this container Allows you to run devices inside the container without the flag. By default, Docker containers are “unprivileged” and cannot, for example, run a Docker daemon inside a Docker container. This is because by default a container is not allowed to access any devices, but a “privileged” container is given access to all devices (see the documentation on cgroups devices).
The flag gives all capabilities to the container. When the operator executes , Docker will enable access to all devices on the host as well as set some configuration in AppArmor or SELinux to allow the container nearly all the same access to the host as processes running outside containers on the host. Additional information about running with is available on the Docker Blog.
If you want to limit access to a specific device or devices you can use the flag. It allows you to specify one or more devices that will be accessible within the container.
By default, the container will be able to , , and these devices. This can be overridden using a third set of options to each flag:
In addition to , the operator can have fine grain control over the capabilities using and . By default, Docker has a default list of capabilities that are kept. The following table lists the Linux capability options which are allowed by default and can be dropped.
Capability Key Capability Description AUDIT_WRITE Write records to kernel auditing log. CHOWN Make arbitrary changes to file UIDs and GIDs (see chown(2)). DAC_OVERRIDE Bypass file read, write, and execute permission checks. FOWNER Bypass permission checks on operations that normally require the file system UID of the process to match the UID of the file. FSETID Don’t clear set-user-ID and set-group-ID permission bits when a file is modified. KILL Bypass permission checks for sending signals. MKNOD Create special files using mknod(2). NET_BIND_SERVICE Bind a socket to internet domain privileged ports (port numbers less than 1024). NET_RAW Use RAW and PACKET sockets. SETFCAP Set file capabilities. SETGID Make arbitrary manipulations of process GIDs and supplementary GID list. SETPCAP Modify process capabilities. SETUID Make arbitrary manipulations of process UIDs. SYS_CHROOT Use chroot(2), change root directory. The next table shows the capabilities which are not granted by default and may be added.
Capability Key Capability Description AUDIT_CONTROL Enable and disable kernel auditing; change auditing filter rules; retrieve auditing status and filtering rules. AUDIT_READ Allow reading the audit log via multicast netlink socket. BLOCK_SUSPEND Allow preventing system suspends. BPF Allow creating BPF maps, loading BPF Type Format (BTF) data, retrieve JITed code of BPF programs, and more. CHECKPOINT_RESTORE Allow checkpoint/restore related operations. Introduced in kernel 5.9. DAC_READ_SEARCH Bypass file read permission checks and directory read and execute permission checks. IPC_LOCK Lock memory (mlock(2), mlockall(2), mmap(2), shmctl(2)). IPC_OWNER Bypass permission checks for operations on System V IPC objects. LEASE Establish leases on arbitrary files (see fcntl(2)). LINUX_IMMUTABLE Set the FS_APPEND_FL and FS_IMMUTABLE_FL i-node flags. MAC_ADMIN Allow MAC configuration or state changes. Implemented for the Smack LSM. MAC_OVERRIDE Override Mandatory Access Control (MAC). Implemented for the Smack Linux Security Module (LSM). NET_ADMIN Perform various network-related operations. NET_BROADCAST Make socket broadcasts, and listen to multicasts. PERFMON Allow system performance and observability privileged operations using perf_events, i915_perf and other kernel subsystems SYS_ADMIN Perform a range of system administration operations. SYS_BOOT Use reboot(2) and kexec_load(2), reboot and load a new kernel for later execution. SYS_MODULE Load and unload kernel modules. SYS_NICE Raise process nice value (nice(2), setpriority(2)) and change the nice value for arbitrary processes. SYS_PACCT Use acct(2), switch process accounting on or off. SYS_PTRACE Trace arbitrary processes using ptrace(2). SYS_RAWIO Perform I/O port operations (iopl(2) and ioperm(2)). SYS_RESOURCE Override resource Limits. SYS_TIME Set system clock (settimeofday(2), stime(2), adjtimex(2)); set real-time (hardware) clock. SYS_TTY_CONFIG Use vhangup(2); employ various privileged ioctl(2) operations on virtual terminals. SYSLOG Perform privileged syslog(2) operations. WAKE_ALARM Trigger something that will wake up the system. Further reference information is available on the capabilities(7) - Linux man page, and in the Linux kernel source code.
Both flags support the value , so to allow a container to use all capabilities except for :
The and flags accept capabilities to be specified with a prefix. The following examples are therefore equivalent:
For interacting with the network stack, instead of using they should use to modify the network interfaces.
To mount a FUSE based filesystem, you need to combine both and :
The default seccomp profile will adjust to the selected capabilities, in order to allow use of facilities allowed by the capabilities, so you should not have to adjust this.
The container can have a different logging driver than the Docker daemon. Use the with the command to configure the container’s logging driver. The following options are supported:
Driver Description Disables any logging for the container. won’t be available with this driver. Logs are stored in a custom format designed for minimal overhead. Default logging driver for Docker. Writes JSON messages to file. No logging options are supported for this driver. Syslog logging driver for Docker. Writes log messages to syslog. Journald logging driver for Docker. Writes log messages to . Graylog Extended Log Format (GELF) logging driver for Docker. Writes log messages to a GELF endpoint likeGraylog or Logstash. Fluentd logging driver for Docker. Writes log messages to (forward input). Amazon CloudWatch Logs logging driver for Docker. Writes log messages to Amazon CloudWatch Logs. Splunk logging driver for Docker. Writes log messages to using Event Http Collector. Event Tracing for Windows (ETW) events. Writes log messages as Event Tracing for Windows (ETW) events. Only Windows platforms. Google Cloud Platform (GCP) Logging. Writes log messages to Google Cloud Platform (GCP) Logging. Rapid7 Logentries. Writes log messages to Rapid7 Logentries. The command is available only for the and logging drivers. For detailed information on working with logging drivers, see Configure logging drivers.
When a developer builds an image from a Dockerfile or when committing it, the developer can set a number of default parameters that take effect when the image starts up as a container.
Four of the Dockerfile commands cannot be overridden at runtime: , , , and . Everything else has a corresponding override in . We’ll go through what the developer might have set in each Dockerfile instruction and how the operator can override that setting.
CMD (Default Command or Options) ENTRYPOINT (Default Command to Execute at Runtime) EXPOSE (Incoming Ports) ENV (Environment Variables) HEALTHCHECK VOLUME (Shared Filesystems) USER WORKDIR Recall the optional in the Docker commandline:
This command is optional because the person who created the may have already provided a default using the Dockerfile instruction. As the operator (the person running a container from the image), you can override that instruction just by specifying a new .
If the image also specifies an then the or get appended as arguments to the .
The of an image is similar to a because it specifies what executable to run when the container starts, but it is (purposely) more difficult to override. The gives a container its default nature or behavior, so that when you set an you can run the container as if it were that binary, complete with default options, and you can pass in more options via the . But, sometimes an operator may want to run something else inside the container, so you can override the default at runtime by using a string to specify the new . Here is an example of how to run a shell in a container that has been set up to automatically run something else (like ):
or two examples of how to pass more parameters to that ENTRYPOINT:
You can reset a containers entrypoint by passing an empty string, for example:
Note
Passing will clear out any default command set on the image (i.e. any instruction in the Dockerfile used to build it).
The following command options work with container networking:
With the exception of the directive, an image developer hasn’t got much control over networking. The instruction defines the initial incoming ports that provide services. These ports are available to processes inside the container. An operator can use the option to add to the exposed ports.
To expose a container’s internal port, an operator can start the container with the or flag. The exposed port is accessible on the host and the ports are available to any client that can reach the host.
The option publishes all the ports to the host interfaces. Docker binds each exposed port to a random port on the host. The range of ports are within an ephemeral port range defined by . Use the flag to explicitly map a single port or range of ports.
The port number inside the container (where the service listens) does not need to match the port number exposed on the outside of the container (where clients connect). For example, inside the container an HTTP service is listening on port 80 (and so the image developer specifies in the Dockerfile). At runtime, the port might be bound to 42800 on the host. To find the mapping between the host ports and the exposed ports, use .
If the operator uses when starting a new client container in the default bridge network, then the client container can access the exposed port via a private networking interface. If is used when starting a container in a user-defined network as described in Networking overview, it will provide a named alias for the container being linked to.
Docker automatically sets some environment variables when creating a Linux container. Docker does not set any environment variables when creating a Windows container.
The following environment variables are set for Linux containers:
Variable Value Set based on the value of The hostname associated with the container Includes popular directories, such as if the container is allocated a pseudo-TTY Additionally, the operator can set any environment variable in the container by using one or more flags, even overriding those mentioned above, or already defined by the developer with a Dockerfile . If the operator names an environment variable without specifying a value, then the current value of the named variable is propagated into the container’s environment:
Similarly the operator can set the HOSTNAME (Linux) or COMPUTERNAME (Windows) with .
Example:
The health status is also displayed in the output.
The example below mounts an empty tmpfs into the container with the , , , and options.
Note
When using systemd to manage the Docker daemon’s start and stop, in the systemd unit file there is an option to control mount propagation for the Docker daemon itself, called . The value of this setting may cause Docker to not see mount propagation changes made on the mount point. For example, if this value is , you may not be able to use the or propagation on a volume.
The volumes commands are complex enough to have their own documentation in section Use volumes. A developer can define one or more ’s associated with an image, but only the operator can give access from one container to another (or from a container to a volume mounted on the host).
The must always be an absolute path such as . The can either be an absolute path or a value. If you supply an absolute path for the , Docker bind-mounts to the path you specify. If you supply a , Docker creates a named volume by that .
A value must start with an alphanumeric character, followed by , (underscore), (period) or (hyphen). An absolute path starts with a (forward slash).
For example, you can specify either or for a value. If you supply the value, Docker creates a bind mount. If you supply the specification, Docker creates a named volume.
(id = 0) is the default user within a container. The image developer can create additional users. Those users are accessible by name. When passing a numeric ID, the user does not have to exist in the container.
The developer can set a default user to run the first process with the Dockerfile instruction. When starting a container, the operator can override the instruction by passing the option.
Note: if you pass a numeric uid, it must be in the range of 0-2147483647. If you pass a username, the user must exist in the container.
The default working directory for running binaries within a container is the root directory (). It is possible to set a different working directory with the Dockerfile command. The operator can override this with: