Rate limiting when publishing subgraph schemas

Other Apollo Topics
, , ,
1

Since the outage caused by AWS, we have been unable to publish our graphql schemas.

When we run rover subgraph schema publish, we get a RATE_LIMIT_EXCEEDED response. This is not intermittent, this happens every time. I have revoked a few API keys, and shut down our processes that publish schema in order to solve any rate limiting issues, but the issues persist. It seems like we are not being rate limited on requests, as we only publish new schemas maybe once every 3 hours. It also seems suspicious that this started happening during the incident.

I have filed a support ticket as well, but have gotten no response. Is there something different we should be doing here? Is this something on Apollo’s side? Any help or response is appreciated.

Edit:

We also were trying to remove a few deactivated users from our organization in the console, and also got a rate limit error. Strange….

2

Alix, Thanks for reaching out- As communicated in the support ticket, we saw 700+ publishes per minute on metronome-v1graph, our rate limit is 400RPM.–> Platform Limits - Apollo GraphQL Docs

Please work with our support Engineer via the support ticket further- we are pushing a hot fix to make an exception. I believe there is a backlog of requests due to the Cloud vendor outage from Monday