0
votes

I'm getting repeated Deploy Errors on two separate projects (same code) just now.

Debug log from 'firebase deploy' looks like:

[debug] >>> HTTP REQUEST GET https://cloudfunctions.googleapis.com/v1beta2/operations/<LONG_ID_HERE>  
 Tue May 16 2017 13:06:06 GMT-0400 (EDT)
[debug] <<< HTTP RESPONSE 200
[debug] [functions] operation poll: <FUNCTION NAME>: update operations/<LONG_ID_HERE> is not done.

This goes on for about 15 minutes and then it dies.

Unfortunately, it appears to leave my functions partially running - the function that it chokes on appears to be non-active after the failed deployment.

Although the code I'm deploying to two firebase projects (dev / staging) is the same code, the deployment seems to choke on different functions.

In this case, both functions are database trigger functions.

The firebase status page appears all-green.

Is there any way to get further information? Could this be my issue - or is this like https://status.firebase.google.com/incident/Functions/17001

1
Both projects deployed successfully after some time and several attempts. The most alarming aspect of this was that even though the deployment failed, the functions were left in a partially inoperable state. This could be problematic in a fully production system, right? - yarrumretep

1 Answers

0
votes

Check the Error Log information if you linked Firebase to Google Cloud Platform. I think the problem here is that there are some entities in your database which do not comply with the non-active function that you have in the codebase. Even if it's a non-active function, make sure that it compiles and executes without any error on your local machine. If you are no longer in need of a function, comment it out totally. Firebase/GCP will continue to compile and launch new instances. When it maxes out the instances, it quits.