How to check what port a pod is listening on with kubectl and not looking at the dockerFile? - kubernetes

Question

I have a pod running and want to port forward so i can access the pod from the internal network.
I don't know what port it is listening on though, there is no service yet.
I describe the pod:
$ kubectl describe pod queue-l7wck
Name: queue-l7wck
Namespace: default
Priority: 0
Node: minikube/192.168.64.3
Start Time: Wed, 18 Dec 2019 05:13:56 +0200
Labels: app=work-queue
chapter=jobs
component=queue
Annotations: <none>
Status: Running
IP: 172.17.0.2
IPs:
IP: 172.17.0.2
Controlled By: ReplicaSet/queue
Containers:
queue:
Container ID: docker://13780475170fa2c0d8e616ba1a3b1554d31f404cc0a597877e790cbf01838e63
Image: gcr.io/kuar-demo/kuard-amd64:blue
Image ID: docker-pullable://gcr.io/kuar-demo/kuard-amd64#sha256:1ecc9fb2c871302fdb57a25e0c076311b7b352b0a9246d442940ca8fb4efe229
Port: <none>
Host Port: <none>
State: Running
Started: Wed, 18 Dec 2019 05:14:02 +0200
Ready: True
Restart Count: 0
Environment: <none>
Mounts:
/var/run/secrets/kubernetes.io/serviceaccount from default-token-mbn5b (ro)
Conditions:
Type Status
Initialized True
Ready True
ContainersReady True
PodScheduled True
Volumes:
default-token-mbn5b:
Type: Secret (a volume populated by a Secret)
SecretName: default-token-mbn5b
Optional: false
QoS Class: BestEffort
Node-Selectors: <none>
Tolerations: node.kubernetes.io/not-ready:NoExecute for 300s
node.kubernetes.io/unreachable:NoExecute for 300s
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Normal Scheduled <unknown> default-scheduler Successfully assigned default/queue-l7wck to minikube
Normal Pulling 31h kubelet, minikube Pulling image "gcr.io/kuar-demo/kuard-amd64:blue"
Normal Pulled 31h kubelet, minikube Successfully pulled image "gcr.io/kuar-demo/kuard-amd64:blue"
Normal Created 31h kubelet, minikube Created container queue
Normal Started 31h kubelet, minikube Started container queue
even the JSON has nothing:
$ kubectl get pods queue-l7wck -o json
{
"apiVersion": "v1",
"kind": "Pod",
"metadata": {
"creationTimestamp": "2019-12-18T03:13:56Z",
"generateName": "queue-",
"labels": {
"app": "work-queue",
"chapter": "jobs",
"component": "queue"
},
"name": "queue-l7wck",
"namespace": "default",
"ownerReferences": [
{
"apiVersion": "apps/v1",
"blockOwnerDeletion": true,
"controller": true,
"kind": "ReplicaSet",
"name": "queue",
"uid": "a9ec07f7-07a3-4462-9ac4-a72226f54556"
}
],
"resourceVersion": "375402",
"selfLink": "/api/v1/namespaces/default/pods/queue-l7wck",
"uid": "af43027d-8377-4227-b366-bcd4940b8709"
},
"spec": {
"containers": [
{
"image": "gcr.io/kuar-demo/kuard-amd64:blue",
"imagePullPolicy": "Always",
"name": "queue",
"resources": {},
"terminationMessagePath": "/dev/termination-log",
"terminationMessagePolicy": "File",
"volumeMounts": [
{
"mountPath": "/var/run/secrets/kubernetes.io/serviceaccount",
"name": "default-token-mbn5b",
"readOnly": true
}
]
}
],
"dnsPolicy": "ClusterFirst",
"enableServiceLinks": true,
"nodeName": "minikube",
"priority": 0,
"restartPolicy": "Always",
"schedulerName": "default-scheduler",
"securityContext": {},
"serviceAccount": "default",
"serviceAccountName": "default",
"terminationGracePeriodSeconds": 30,
"tolerations": [
{
"effect": "NoExecute",
"key": "node.kubernetes.io/not-ready",
"operator": "Exists",
"tolerationSeconds": 300
},
{
"effect": "NoExecute",
"key": "node.kubernetes.io/unreachable",
"operator": "Exists",
"tolerationSeconds": 300
}
],
"volumes": [
{
"name": "default-token-mbn5b",
"secret": {
"defaultMode": 420,
"secretName": "default-token-mbn5b"
}
}
]
},
"status": {
"conditions": [
{
"lastProbeTime": null,
"lastTransitionTime": "2019-12-18T03:13:56Z",
"status": "True",
"type": "Initialized"
},
{
"lastProbeTime": null,
"lastTransitionTime": "2019-12-18T03:14:02Z",
"status": "True",
"type": "Ready"
},
{
"lastProbeTime": null,
"lastTransitionTime": "2019-12-18T03:14:02Z",
"status": "True",
"type": "ContainersReady"
},
{
"lastProbeTime": null,
"lastTransitionTime": "2019-12-18T03:13:56Z",
"status": "True",
"type": "PodScheduled"
}
],
"containerStatuses": [
{
"containerID": "docker://13780475170fa2c0d8e616ba1a3b1554d31f404cc0a597877e790cbf01838e63",
"image": "gcr.io/kuar-demo/kuard-amd64:blue",
"imageID": "docker-pullable://gcr.io/kuar-demo/kuard-amd64#sha256:1ecc9fb2c871302fdb57a25e0c076311b7b352b0a9246d442940ca8fb4efe229",
"lastState": {},
"name": "queue",
"ready": true,
"restartCount": 0,
"started": true,
"state": {
"running": {
"startedAt": "2019-12-18T03:14:02Z"
}
}
}
],
"hostIP": "192.168.64.3",
"phase": "Running",
"podIP": "172.17.0.2",
"podIPs": [
{
"ip": "172.17.0.2"
}
],
"qosClass": "BestEffort",
"startTime": "2019-12-18T03:13:56Z"
}
}
How do you checker what port a pod is listening on with kubectl?
Update
If I ssh into the pod and run netstat -tulpn as suggested in the comments I get:
$ kubectl exec -it queue-pfmq2 -- sh
~ $ netstat -tulpn
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name
tcp 0 0 :::8080 :::* LISTEN 1/kuard
But this method is not using kubectl.

Answer 1

Your container image has a port opened during the build (looks like port 8080 in your case) using the EXPOSE command in the Dockerfile. Since the exposed port is baked into the image, k8s does not keep track of this open port since k8s does not need to take steps to open it.
Since k8s is not responsible for opening the port, you won't be able to find the listening port using kubectl or checking the pod YAML

Answer 2

Try the combination of both kubectl and your Linux command to get the Port container is listening on:
kubectl exec <pod name here> -- netstat -tulpn
Further you can pipe this result with grep to narrow the findings if required eg.
kubectl exec <pod name here> -- netstat -tulpn | grep "search string"
Note: It will work only if your container's base image supports the command netstat. and as per your Update section it seems it supports.
Above solution is nothing but a smart use of the commands you have used in two parts first to exec the container in interactive mode using -it second in the container to list the listening port.

Answer 3

One answer suggested to run netstat inside the container.
This only works if netstat is part of the container's image.
As an alternative, you can run netstat on the host executing it in the container's network namespace..
Get the container's process ID on the host (this is the application running inside the container). Then change to the container's network namespace (run as root on the host):
host# PS1='container# ' nsenter -t <PID> -n
Modifying the PS1 environment variable is used to show a different prompt while you are in the container's network namespace.
Get the listening ports in the container:
container# netstat -na
....
container# exit

Answer 4

If who created the image added the right Openshift label then you can use the following command (unfortunately your image does not have the label) :
skopeo inspect docker://image-url:tag | grep expose-service
e.g.
skopeo inspect docker://quay.io/redhattraining/loadtest:v1.0 | grep expose-service
output:
"io.openshift.expose-services": "8080:http"
So 8080 is the port exposed by the image
Hope this helps

Answer 5

normally. a container will able to run curl . so you can use curl to check whether a port is open.
for port in 8080 50000 443 8443;do curl -I - connect-timeout 1 127.0.0.1:$port;done
this can be run with sh.