Description of problem: For a Node.js app, although it is by design that hot deployment won't take effect until another deployment after supervisor process is started (that is to say, the node process will be restarted when we push hot_deploy marker to the app's repo), it still prints "remote: Not stopping cartridge nodejs because hot deploy is enabled" the first time we push hot_deploy marker while the node process is actually restarted. It would be more user friendly if we could keep users informed of this design. Version-Release number of selected component (if applicable): devenv_3628 How reproducible: always Steps to Reproduce: 1. create a node.js application 2. SSH into this app, and record the pid in its nodejs/run/node.pid 3. create a hot deploy marker 4. push the marker to the app's repo 5. check the pid file again to see if the pid changes Actual results: When pushing the hot_deploy marker, the node.js app was actually restarted (node.pid changed), but the messages printed in console were misleading: remote: Not stopping cartridge nodejs because hot deploy is enabled ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ remote: Saving away previously installed Node modules remote: npm info it worked if it ends with ok remote: npm info using npm.37 remote: npm info using node.20 remote: npm info preinstall OpenShift-Sample-App.0 remote: npm info build /var/lib/openshift/627c4dacfffc11e287d022000aa416e1/app-root/runtime/repo remote: npm info linkStuff OpenShift-Sample-App.0 remote: npm info install OpenShift-Sample-App.0 remote: npm info postinstall OpenShift-Sample-App.0 remote: npm info ok remote: Starting application njs remote: Not starting cartridge nodejs because hot deploy is enabled ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Expected results: Notify Node.js users with correct information the first time pushing hot_deploy marker. Additional info:
https://github.com/openshift/origin-server/pull/3334
Commit pushed to master at https://github.com/openshift/origin-server https://github.com/openshift/origin-server/commit/8880122cc74f2ed9794460863e5dd53bde530165 Bug 994922 Warn user of supervisor control when we detect 'hot_deploy' marker. When the hot_deploy marker is present, stop nor start is invoked, so we need to check the presence of the marker in a different way. Notice that the first time we push 'hot_deploy' marker, the stop cycle will not warn that it detected the marker (because it doesn't exist on the gear).
Verified in devenv_3647: Upon the first time we push 'hot_deploy' marker, it prompts: remote: Not starting cartridge nodejs because hot deploy is enabled remote: hot_deploy marker found. Switching to using supervisor. Subsequent pushes won't require restart until the marker is removed. To ssh://e65412cc04ac11e3953622000aa42b7a.rhcloud.com/~/git/njs.git/ 037cb0d..835966d master -> master which is desirable. Thanks, Zhe Wang