I'm getting a weird error. I have an ApolloApi and I'm getting a 503 response, from a specific query, however, the lambda executes without an error and returns the given values from the db. (I can see it in my logs). There is something wrong with the lambda communicating with API Gateway. Anyone have a guess what it could be? This is only happening with a specific graphql query, other queries are getting 200 as expected.

Can you log the exact output from the lambda before returning it?

Is it taking too long to return? APIG has a 30 second timeout

I set the timeout to 260, I'm able to log the data that is to be returned

The api gateway is hardcoded AFAIK, can’t be changed, you can check the timestamps on the START and END events on the cloudwatch logs