hello-cf

This simple application displays the request URL, the port number an some application meta-data for the deployed application instance. Displaying the port number supports a demonstration where, when deployed to Cloud Foundry and the application crashes, a instance will be created and the app will be redeployed and started.

The script for the demo is as follows:

Clone this repo

Clone the repo to your local computer:

git clone https://github.com/christofmarti/hello-cf.git

and change to the application directory:

cd hello-cf

Deploy the app

Deploy your app instances using cf push. Subsititute <your-unique-app-id> with a unique name (e.g. your group id of the course). The name (-n) of your application has to be unique within the same dns <domain> of the plattform (cfapps.io in case of PWS). hello-cf is your (internal) name for the app, which has to be provided for each cf-command.

cf push hello-cf -m 128M -n hello-cf-<your-unique-app-id> 

This will deploy the application to http://hello-cf-<your-unique-app-id>.<domain> where <domain is the default domain of the plattform (in case of PWS this is cfapps.io).

Start tailing the logs

Recommend splitting your terminal and in one half start logging with:

cf logs hello-cf

Access the app in your browser

Check out the environment information which is provided about the application.

http://hello-cf-<your-unique-app-id>.<domain>/broken

where <your-unique-app-id> is the name you've chosen above and <domain> is the DNS domain you deployed your app to (in case of PWS this is usually cfapps.io).

Scale your application

Scale your application to 3 instances - this will allow you to show that in the short time one instance is crashed, the others are still serving traffic.

cf scale hello-cf -i 3

Access the app in your browser once more

Hit refresh a few times and point out the two different port numbers displayed. Have audience remember the two port numbers

Kill one of the app instances

You can do this by accessing the url

http://hello-cf-<your-unique-app-id>.<domain>/broken

Look at log and app instances

You'll have to be quick - do a

cf app hello-cf

showing that one of the app instances is down. Highlight stack trace in the logs and then run this again showing the app is restarted:

cf app hello-cf

Refresh app in browser

Refresh until you show that one of the port numbers remains the same but one has been replaced with a new port number.

Reminder!!! Make sure you go back to the URL without the "/broken" suffix or you'll crash your app again.