Add a Docker multi-node setup.
This commit is contained in:
58
docs/getting-started-guides/docker-multinode/testing.md
Normal file
58
docs/getting-started-guides/docker-multinode/testing.md
Normal file
@@ -0,0 +1,58 @@
|
||||
## Testing your Kubernetes cluster.
|
||||
|
||||
To validate that your node(s) have been added, run:
|
||||
|
||||
```sh
|
||||
kubectl get nodes
|
||||
```
|
||||
|
||||
That should show something like:
|
||||
```
|
||||
NAME LABELS STATUS
|
||||
10.240.99.26 <none> Ready
|
||||
127.0.0.1 <none> Ready
|
||||
```
|
||||
|
||||
If the status of any node is ```Unknown``` or ```NotReady``` your cluster is broken, double check that all containers are running properly, and if all else fails, contact us on IRC at
|
||||
```#google-containers``` for advice.
|
||||
|
||||
### Run an application
|
||||
```sh
|
||||
kubectl -s http://localhost:8080 run-container nginx --image=nginx --port=80
|
||||
```
|
||||
|
||||
now run ```docker ps``` you should see nginx running. You may need to wait a few minutes for the image to get pulled.
|
||||
|
||||
### Expose it as a service:
|
||||
```sh
|
||||
kubectl expose rc nginx --port=80
|
||||
```
|
||||
|
||||
This should print:
|
||||
```
|
||||
NAME LABELS SELECTOR IP PORT(S)
|
||||
nginx <none> run-container=nginx <ip-addr> 80/TCP
|
||||
```
|
||||
|
||||
Hit the webserver:
|
||||
```sh
|
||||
curl <insert-ip-from-above-here>
|
||||
```
|
||||
|
||||
Note that you will need run this curl command on your boot2docker VM if you are running on OS X.
|
||||
|
||||
### Scaling
|
||||
|
||||
Now try to scale up the nginx you created before:
|
||||
|
||||
```sh
|
||||
kubectl resize rc nginx --replicas=3
|
||||
```
|
||||
|
||||
And list the pods
|
||||
|
||||
```sh
|
||||
kubectl get pods
|
||||
```
|
||||
|
||||
You should see pods landing on the newly added machine.
|
Reference in New Issue
Block a user