First, confirm that no Elasticsearch containers are currently running on our machine:

$ docker ps -a \
 --filter "name=elasticsearch" \
 --format "table {{.ID}}\t{{.Image}}\t{{.Command}}\t{{.Names}}"
CONTAINER ID IMAGE COMMAND NAMES

We can now use kubectl run to run an image inside a Pod, and deploy it onto our cluster:

$ kubectl run elasticsearch --image=docker.elastic.co/elasticsearch/elasticsearch-oss:6.3.2 --port=9200 --port=9300
deployment.apps "elasticsearch" created

Now, when we check the Pods that have been deployed onto our cluster, we can see a new elasticsearch-656d7c98c6-s6v58 Pod:

$ kubectl get pods
NAME READY STATUS RESTARTS AGE
elasticsearch-656d7c98c6-s6v58 0/1 ContainerCreating 0 9s

It may take some time for the Pod to initiate, especially if the Docker image is not available locally and needs to be downloaded. Eventually, you should see the READY value become 1/1:

$ kubectl get pods
NAME READY STATUS RESTARTS AGE
elasticsearch-656d7c98c6-s6v58 1/1 Running 0 1m