Deploy on Kubernetes
Estimated reading time: 3 minutes
Docker Desktop includes a standalone Kubernetes server and client, as well as Docker CLI integration. The Kubernetes server runs locally within your Docker instance, is not configurable, and is a single-node cluster.
The Kubernetes server runs within a Docker container on your local system, and is only for local testing. When Kubernetes support is enabled, you can deploy your workloads, in parallel, on Kubernetes, Swarm, and as standalone containers. Enabling or disabling the Kubernetes server does not affect your other workloads.
See Docker Desktop for Mac > Getting started to enable Kubernetes and begin testing the deployment of your workloads on Kubernetes.
Use Docker commands
To ensure that the docker stack commands are available on your installation, try
docker stack ls
. If you see the message failed to find a Stack API version
then the commands described on this page are not available. This can be confirmed
with docker version
which should contain the line StackAPI: Unknown
.
If docker stack commands are working then you can deploy a stack on Kubernetes with
docker stack deploy
, the docker-compose.yml
file, and the name of the stack.
docker stack deploy --compose-file /path/to/docker-compose.yml mystack
docker stack services mystack
You can see the service deployed with the kubectl get services
command.
Specify a namespace
By default, the default
namespace is used. You can specify a namespace with
the --namespace
flag.
docker stack deploy --namespace my-app --compose-file /path/to/docker-compose.yml mystack
Run kubectl get services -n my-app
to see only the services deployed in the
my-app
namespace.
Override the default orchestrator
While testing Kubernetes, you may want to deploy some workloads in swarm mode.
Use the DOCKER_STACK_ORCHESTRATOR
variable to override the default orchestrator for
a given terminal session or a single Docker command. This variable can be unset
(the default, in which case Kubernetes is the orchestrator) or set to swarm
or
kubernetes
. The following command overrides the orchestrator for a single
deployment, by setting the variable
at the start of the command itself.
DOCKER_STACK_ORCHESTRATOR=swarm docker stack deploy --compose-file /path/to/docker-compose.yml mystack
Alternatively, the --orchestrator
flag may be set to swarm
or kubernetes
when deploying to override the default orchestrator for that deployment.
docker stack deploy --orchestrator swarm --compose-file /path/to/docker-compose.yml mystack
Note
Deploying the same app in Kubernetes and swarm mode may lead to conflicts with ports and service names.
Use the kubectl command
The mac Kubernetes integration provides the Kubernetes CLI command
at /usr/local/bin/kubectl
. This location may not be in your shell’s PATH
variable, so you may need to type the full path of the command or add it to
the PATH
. For more information about kubectl
, see the
official kubectl
documentation.
You can test the command by listing the available nodes:
kubectl get nodes
NAME STATUS ROLES AGE VERSION
docker-desktop Ready master 3h v1.8.2
Example app
Docker has created the following demo app that you can deploy to swarm mode or
to Kubernetes using the docker stack deploy
command.
version: "3.9"
services:
web:
image: dockersamples/k8s-wordsmith-web
ports:
- "80:80"
words:
image: dockersamples/k8s-wordsmith-api
deploy:
replicas: 5
endpoint_mode: dnsrr
resources:
limits:
memory: 50M
reservations:
memory: 50M
db:
image: dockersamples/k8s-wordsmith-db
If you already have a Kubernetes YAML file, you can deploy it using the
kubectl
command.