docker service create

Create a new service Docker CLI reference

75 options Its page in the Docker CLI reference

Options

75 in the Docker CLI reference

Show its 75 options

--cap-addAdd Linux capabilities

--cap-dropDrop Linux capabilities

--configSpecify configurations to expose to the service

--constraintPlacement constraints

--container-labelContainer labels

--credential-specCredential spec for managed service account (Windows only)

--detach, -dExit immediately instead of waiting for the service to converge

--dns-optionSet DNS options

--dns-searchSet custom DNS search domains

--dnsSet custom DNS servers

--endpoint-modeEndpoint mode (vip or dnsrr)

--entrypointOverwrite the default ENTRYPOINT of the image

--env, -eSet environment variables

--env-fileRead in a file of environment variables

--generic-resourceUser defined resources

--groupSet one or more supplementary user groups for the container

--health-cmdCommand to run to check health

--health-intervalTime between running the check (ms\

--health-retriesConsecutive failures needed to report unhealthy

--health-start-intervalTime between running the check during the start period (ms\

--health-start-periodStart period for the container to initialize before counting retries towards unstable (ms\

--health-timeoutMaximum time to allow one check to run (ms\

--hostSet one or more custom host-to-IP mappings (host:ip)

--hostnameContainer hostname

--initUse an init inside each service container to forward signals and reap processes

--isolationService container isolation mode

--label, -lService labels

--limit-cpuLimit CPUs

--limit-memoryLimit Memory

--limit-pidsLimit maximum number of processes (default 0 = unlimited)

--log-driverLogging driver for service

--log-optLogging driver options

--max-concurrentNumber of job tasks to run concurrently (default equal to --replicas)

--memory-swapSwap Bytes (-1 for unlimited)

--memory-swappinessTune memory swappiness (0-100), -1 to reset to default

--modeService mode (replicated, global, replicated-job, global-job)

--mountAttach a filesystem mount to the service

--nameService name

--networkNetwork attachments

--no-healthcheckDisable any container-specified HEALTHCHECK

--no-resolve-imageDo not query the registry to resolve image digest and supported platforms

--oom-score-adjTune host's OOM preferences (-1000 to 1000)

--placement-prefAdd a placement preference

--publish, -pPublish a port as a node port

--quiet, -qSuppress progress output

--read-onlyMount the container's root filesystem as read only

--replicas-max-per-nodeMaximum number of tasks per node (default 0 = unlimited)

--replicasNumber of tasks

--reserve-cpuReserve CPUs

--reserve-memoryReserve Memory

--restart-conditionRestart when condition is met (none, on-failure, any) (default any)

--restart-delayDelay between restart attempts (ns\

--restart-max-attemptsMaximum number of restarts before giving up

--restart-windowWindow used to evaluate the restart policy (ns\

--rollback-delayDelay between task rollbacks (ns\

--rollback-failure-actionAction on rollback failure (pause, continue) (default pause)

--rollback-max-failure-ratioFailure rate to tolerate during a rollback (default 0)

--rollback-monitorDuration after each task rollback to monitor for failure (ns\

--rollback-orderRollback order (start-first, stop-first) (default stop-first)

--rollback-parallelismMaximum number of tasks rolled back simultaneously (0 to roll back all at once)

--secretSpecify secrets to expose to the service

--stop-grace-periodTime to wait before force killing a container (ns\

--stop-signalSignal to stop the container

--sysctlSysctl options

--tty, -tAllocate a pseudo-TTY

--ulimitUlimit options

--update-delayDelay between updates (ns\

--update-failure-actionAction on update failure (pause, continue, rollback) (default pause)

--update-max-failure-ratioFailure rate to tolerate during an update (default 0)

--update-monitorDuration after each task update to monitor for failure (ns\

--update-orderUpdate order (start-first, stop-first) (default stop-first)

--update-parallelismMaximum number of tasks updated simultaneously (0 to update all at once)

--user, -uUsername or UID (format: <name\

--with-registry-authSend registry authentication details to swarm agents

--workdir, -wWorking directory inside the container