We have a situation where we are deploying an app to two different Cloud Foundry API endpoints using TravisCI in a continuous delivery approach to production. It is working well if we push the updated app using a serial approach, to one API endpoint and then the other after it succeeds. But when we try to push to both in parallel with fancy login/logout using Cloud Foundry CLI it fails around 1/2 of the time. This seems to be due to the shared state between shell processes being run by the same user. Is anyone else push an app to multiple Cloud Foundry API endpoints in parallel and have an approach that works more effectively. We'd like to reduce the deployment timespan if we can. Thanks.- Chris
--
You received this message because you are subscribed to the Google Groups "Cloud Foundry Developers" group.
To view this discussion on the web visit https://groups.google.com/a/cloudfoundry.org/d/msgid/vcap-dev/f5e5dbac-07f2-4764-aa0b-aa7d3ed912e4%40cloudfoundry.org.
To unsubscribe from this group and stop receiving emails from it, send an email to vcap-dev+u...@cloudfoundry.org.