nc-pod

A simple Kubernetes Pod that runs netcat (nc) typically in the role of the listening endpoint.

The goal of this simple little Pod is to enable any number of Kubernetes networking configuration test scenarios. From a simple intra-pod connectivity test (show here), to any form of advanced networking configuration you can apply to Kubernetes Pods and more specifially, Services.

Building the Docker Image

Build and push the docker image, replacing the Samsung private Quay registry with your target registry.

$ docker build --rm --tag quay.io/samsung_cnct/nc-pod .
$ docker push quay.io/samsung_cnct/nc-pod:latest

Helm Chart

This project might also end up being packaged as a Helm Chart eventually.
However, it's pretty dead simple as is, and might not need to ever become a chart... we'll see.

Configuration

Configuring 'nc' execution is handled by a single environment variable that captures the entire set of 'nc' command line arguments, but not the 'nc' command itself.

For a complete description of 'nc' command line arguments see: man nc(1)

Environment Variable:

NC_CMD_ARGS: Required, a quoted string of nc command line arguments.

Example (excerpt from Deployment spec yaml):

    spec:
      containers:
      - name: nc-pod
        image: quay.io/samsung_cnct/nc-pod
        env:
        - name: NC_CMD_ARGS
          value: "-lk -n -u -p 26500"

Deployment

Using the example nc-pod.yaml file's default values, the project is easily deployed as shown here:

$ kubectl create -f nc-pod.yaml
deployment "nc-pod" created

$ kubectl expose -f nc-pod.yaml 
service "nc-pod" exposed

$ kubectl get po,ep -l app=nc-pod
NAME                         READY     STATUS    RESTARTS   AGE
po/nc-pod-2355685321-3gd8r   1/1       Running   0          58s

NAME        ENDPOINTS          AGE
ep/nc-pod   10.64.0.23:26500   48s

Running / Testing with netcat

From the example Deployment above we tail the output of the created pod.

$ kubectl logs -f nc-pod-2355685321-3gd8r
cmd: /bin/nc -lk -n -u -p 26500

We'll check back on the output of the 'logs' command after the next steps.

Since the service created by the default example isn't exposed via a public IP address, we need to login to either a cluster node or pod. Let's stay away from using the node, and stick to a pod.

$ kubectl run -i -t alps --image=alpine:latest --restart=Never --command /bin/ash
Waiting for pod default/alps to be running, status is Pending, pod ready: false
If you don't see a command prompt, try pressing enter.
/ # 
/ # nc -n -u 10.64.0.23 26500
Example of running netcat in a pod.
^Cpunt!

/ # exit
$

As soon as we hit ruturn on our input line above, we should have seen the following output scroll in our 'logs -f' terminal: