Blueprint sync fails with no error messages

This is my first deployment with a blueprint and it just fails with no error messages. What should I do?

Here is my blueprint:

  # A Docker web service
  - type: web
    name: test-api
    env: docker
    dockerfilePath: ./Dockerfile.production
    dockerContext: .
    region: oregon
    plan: free
    branch: master # optional (uses repo default)
    numInstances: 1 # optional (defaults to 1)
    healthCheckPath: /health
          name: test_db
          property: user
          name: test_db
          property: password
      - key: DATABASE_NAME
          name: test_db
          property: database
      - key: DATABASE_HOST
          name: test_db
          property: host
      - key: AWS_ACCESS_KEY_ID
        generateValue: true
        generateValue: true
        value: "ap-northeast-1"
      - key: AWS_CDN_HOSTNAME
        value: ""
    autoDeploy: true # optional

  - name: test-db
    databaseName: test_production
    ipAllowList: [] # only allow internal connections

For anyone else reading this. I ended up having to manually create a service first so I can have access to the logs then syncing it with the blueprint. If you create a blueprint without any services prior and it has an error, you won’t be able to find logs for it. This seems like a bug with Render.

You shouldn’t have to do that - we have an effort in place to improve error messaging around Blueprints because they can be a bit hidden. What was the error in the end with your Blueprint?

John B

I couldn’t find out what the blueprint error was. There’s no log output. I would have expected a log output for the blueprint itself. Moreover, the resources couldn’t create so I couldn’t even go to the individual resource to see what the log outputs were.

Error output for Blueprints is high on our list to improve this year!

John B

1 Like

We’re testing out since Heroku’s gone down the drain, and I’m just getting a red X next to Create web service api when pushing a render.yaml. There doesn’t appear to be any logs or any messages, so I have no idea why it’s failing. Is there no workaround for this? Kind of a major dealbreaker if we can’t even test it…

I’d also like to add that this is frustrating. Sometimes the sync can highlight what it thinks failed in the render.yaml, but it seems generally it just says: there were problems, here’s your yaml, good luck!

I managed to get a working configuration after a bunch of trial and error, but it would be really useful to:

  1. Have error output that at least shows an underlying error message so we have a hint as to what should change.
  2. Iterate configuration without sending a string of possibly broken commits to a repository. Being able to edit a failed sync in place would be much less annoying than pushing over and over.

Thanks - I’m really thrilled about the promise of this feature. It’s just hard to get up and running!