Automating Deployment Failure Handling


Steve Amerige
 

When I do commands such as bosh deploy, cf push, or cf start, I sometimes see failures that are due to problems in the underlying OpenStack. When I do the commands again, the identical same deploy, push, or start succeeds.

When writing deployment scripts, I'd like to be able to headlessly detect problems and attempt to deal with them before giving up the deployment. Is it possible to get JSON results back from doing BOSH or Cloud Foundry deployment commands such as those above? This might give me more ability to have deployment scripts either decide to just try the deployment again, or to take some specific remedial action depending on what the JSON data reveals.

Any thoughts as to what is available now so that I can do a better job of handling at least some classes of failures in scripts?

Join cf-dev@lists.cloudfoundry.org to automatically receive all group messages.