Skip to content

Latest commit

 

History

History
80 lines (56 loc) · 2 KB

README.md

File metadata and controls

80 lines (56 loc) · 2 KB
Raw

Installation

Inside the app repository, run the following commands:

  cf push test-app
  cf bind-service test-app my_rabbitmq_service
  cf restage test-app

Development

Install dependencies and run all the specs:

bundle --local
bundle exec rspec

The Guard runner is provided for faster feedback from the specs:

bundle exec guard

Usage

To write messages to the queue:

http://test-app.<YOUR_DOMAIN>/write

To read messages from the queue:

http://test-app.<YOUR_DOMAIN>/read

While the write and read endpoints will open a websocket, the store endpoint allows you to write and read single messages using HTTP POST and GET requests. For example:

curl -XPOST -d 'test' http://test-app.<YOUR-DOMAIN>/store

curl -XGET http://test-app.<YOUR-DOMAIN>/store

The queues endpoint allows writing and reading single messages to and from a specific queue using HTTP POST and GET requests:

curl -XPOST -d 'test' http://test-app.<YOUR-DOMAIN>/queues/<YOUR-QUEUE-NAME>

curl -XGET http://test-app.<YOUR-DOMAIN>/queues/<YOUR-QUEUE-NAME>

If you haveHAProxy allows HTTPS traffic only enabled in CloudFoundry, then

  • Replace http with https with http in the above mentioned URL's.
  • Ensure you have a rule on your LoadBalancer (or HAProxy), which forwards secure TCP from port 4443 to TCP on port 80.

Demonstration

  • Open the URLs to read and write in separate browser tabs, side by side.
  • You will see messages being sent & received
  • Identify the HAProxy that the writing app is connected to from the IP in the bold color coded URI string
  • Using BOSH you can bosh stop rabbitmq-haproxy-partition-default_az_guid <index>
  • Observe the writing application fail and then automatically reconnect and continue sending messages
  • You could also bosh stop rabbitmq-server-partition-default_az_guid <index> to simulate a RabbitMQ node failing

License

Released under the Apache 2.0 license

(c) 2015, Pivotal Software