Quantcast
Channel: SCN : Discussion List - SAP HANA Cloud Platform Developer Center
Viewing all articles
Browse latest Browse all 3131

Unable to recover broken app on hana cloud

$
0
0

Hi,

 

it seems that one of our team members has broken our app deployed on the hana cloud and we're unable to recover to a clean state. Let me give some background first. We have a nightly build job which automatically deploys our app using the hana command line tools (always the latest version so currently 1.25.4.3). The job first stops the app, then deploys the new app and finally starts the app again. Now our team member also deployed a development version of our app from his computer to the same app in the cloud. I don't know whether he missed some step from our deployment job or what deployment properties he used but somehow our deployed app is broken now. It always shows up running with 2 processes when started despite being deployed with an elasticity data of [1..1]. One of the processes is always in the starting state but never reaches start and also can not be stopped. When calling stop on the cloud cockpit we get the stopping message but only one process stops and the other one is staying in the starting state. When using the command line tools to stop the app the command finishes without any errors but in the cloud cockpit there is still one process shown in the starting state. In the hana cloud documentation (https://help.hana.ondemand.com/help/frameset.htm?7612f03c711e1014839a8273b0e91070.html) the notes about starting and stopping say

When you stop an application, all application processes are stopped, irrespective of the tool used

but this doesn't seem to be the case for us. Even an undeploy (which is only possible via command line tools since in the cloud cockpit the delete button is never shown because of the starting process) didn't help - the second process persists. Just for fun I also tried to deploy with elasticity data of [3..3] but in this case the app didn't start because 1 process was already in use. In the vague hope that this will resolve our problem I deployed with a vm argument of -Dosgi.clean but that also did not help. I assume the undeploy doesn't remove all state - maybe the command line tools can get a parameter to force the removal of all state? For now it seems like we have to either live with the 2nd process or switch over to another cloud app. I would appreciate any help.

 

Best regards,

Tobias


Viewing all articles
Browse latest Browse all 3131

Trending Articles



<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>