Kong: A server description and deployment testing tool from Eric Holscher and World Online

November 18, 2009 • django, programming • 296 words • about a 1 minute read

A really neat project out of World Online (the company which gave birth to Django itself) that helps you—as a developer—perform browser acceptance tests against a large number of sites and front-ends.

What’s unclear, I think, is how the testing interacts with the deployment process.

Is the code deployed using a traditional CI server?
Can a CI server be configured not to not-/deploy a build if Kong tests fail?
Does Kong handle the actual deployment of the code?
Is there any kind of way to automatically run tests or is it all manual?
If automatic, are there any facilities for notification on failed tests (email, et cetera?)

I’ll probably play around with Kong a little on an existing project, but I’m unsure whether or not it will be useful on project that don’t have a ton of sites or front-ends.

Update: After re-reading, I now understand that Kong is a “deployment testing” tool, not a “server description, deployment, and testing tool.” One would deploy their sites, and then test them using this tool. Seems like it could result in some bad deploys, but if you’ve already got good Django tests, I presume you probably wouldn’t run into too many.

Update again: I fail at reading:

We currently have this wired up to a cron job that runs every 10 minutes. If you set the KONG_MAIL_MANAGERS settings to True, it will send an email to the site managers on a test failure. At some point in the future, I will be integrating Kong into Nagios, so that Nagios will handle the running and alerting of errors. That is eventually the way that it will be run.