Zero Downtime Deployments with Bluemix

In Cloud Foundry (the open source technology behind Bluemix), when you do a cf push, Cloud Foundry will actually stop your app and restart it with the new code that you just uploaded. This presents an issue for a production app or any app that is actually serving users. There is a shortcoming with the current DEA (the part in Cloud Foundry that runs your app) but the next version of the DEA (Diego) will help address this. In the meantime you can do a little scripting to get around this.

The basic flow is as follows:

App A is running (prod)
Deploy App B
Do some tests against App B
Map prod route to App B
Unmap prod route from App A
Stop App A
Delete App A

Here is some starter shell code to do it:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
# deploy.sh

#!/bin/bash
cf unmap-route blue-app mybluemix.net -n cf-blue-green          # make the app unavailable to requests
cf push blue-app

# wait for the blue app to start
while true; do
RESP=`curl -sIL -w "%{http_code}" "blue-app.mybluemix.net" -o /dev/null`
if [[ $RESP == "200" ]]
then break
else sleep 3 &amp;&amp; echo "Waiting for 200 response"
fi
done

# make the blue app available to the router
cf map-route blue-app mybluemix.net -n cf-blue-green

# deploy to the green app
cf unmap-route green-app mybluemix.net -n cf-blue-green
cf push green-app
cf app green-app
cf map-route green-app mybluemix.net -n cf-blue-green