Fly app does not rollback after a failed deploy

Previously, if my deployment failed to pass health checks, the machine would be rolled back to the previous release automatically. That is no longer happening. Now I get this output:

Updating existing machines in 'MY_APP' with rolling strategy
  [1/1] Waiting for MY_ID [app] to have state: started
  Machine MY_ID [app] has state: started
  [1/1] Checking that MY_ID [app] is up and running
  [1/1] Waiting for MY_ID [app] to become healthy: 0/1

WARNING The app is not listening on the expected address and will not be reachable by fly-proxy.
You can fix this by configuring your app to listen on the following addresses:
  - 0.0.0.0:1234
Found these processes inside the machine with open listening sockets:
  PROCESS        | ADDRESSES                              
-----------------*----------------------------------------
  /.fly/hallpass | [IPV6_ADDR]:22  

Error: 
timeout reached waiting for healthchecks to pass for machine MY_ID failed to get VM MY_ID: Get "https://api.machines.dev/v1/apps/MY_APP/machines/MY_ID": net/http: request canceled
note: you can change this timeout with the --wait-timeout flag
Error: Process completed with exit code 1.

And the machine is not rolled back. It just stays down.

Is there somethign I can configure to go back to the old behavior?

Someone asked about this previously but got no response: Health checks and rollback

1 Like

Hi @Adam2! To get similar behavior with Apps v2, you might consider using the canary or bluegreen deployment strategies with fly deploy.

1 Like

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.