search

servicecatalog / oscm-app-shell

A script based framework adapter integrating OSCM with any type of cloud.
Apache License 2.0
1 stars 5 forks source link

Repeated ShellPoolException after restarting server #76

Closed GoebelL closed 4 years ago

GoebelL commented 5 years ago

Version Info OSCM 18.0

Describe the bug The shell adapter manages shell connections in a non-persistent shell pool. After restart all connections are lost, while APP timer keeps polling them. The message org.oscm.app.statemachine.api.StateMachineException: NO_SHELL_FOR_CALLERID gets logged repeatedly every 15 minutes. The customer finds his respective subscriptions staying 'pending update' for ever.

How to Reproduce Steps to reproduce the behavior:

  1. Deploy shell adapter with scripts including the status script.
  2. Subscribe a shell based service and ensure shell scripts are running
  3. Modify the subscription to get in pending update
  4. Restart oscm-app
  5. Trace messages from shell adapter docker logs -f oscm-app

Observed behavior

  1. Repeating logs org.oscm.app.statemachine.api.StateMachineException: NO_SHELL_FOR_CALLERID
  2. APP timer keeps runnung

Expected behavior APP receives exception from querying the instance status and stops the timer for polling the instance. Subscription modification is aborted and rolledback with error message in marketplace UI. Subscription is back to active. Subsequent modifications will lead to new connections.

Screenshots None

Additional context Confidential.

Reported by @uhu133

GoebelL commented 5 years ago

Curent status: The issue cannot be reproduced. "NO_SHELL_FOR_CALLERID gets logged repeatedly every 15 minutes" -> We fail to create such scenario. With or without error to APP, the instance is always completing and stopping the timer thread. Additional: I just received the info, that in preceeding environement there was a mail server issue. As mail notification is mandatory in OSCM, it was part of the cause. After fixing the mail server this issue was no longer observed.