Not sure why you're making up stuff still and cherry picking constants to try and pretend you're right.
Those rates are BEST CASE. I linked the code directly to you that actually does it in your previous comment.
They queue 100[0] users every 5[1] seconds to pull their status, they then update the next check timestamp in the db to be at now + the constant you quote[2] which they use for rate limiting, so at most once per minute.
So unless they have under 1,000 users, then it won't ever be "every single minute."
Those rates are BEST CASE. I linked the code directly to you that actually does it in your previous comment.
They queue 100[0] users every 5[1] seconds to pull their status, they then update the next check timestamp in the db to be at now + the constant you quote[2] which they use for rate limiting, so at most once per minute.
So unless they have under 1,000 users, then it won't ever be "every single minute."
[0]: https://github.com/christianselig/apollo-backend/blob/b992d2...
[1]: https://github.com/christianselig/apollo-backend/blob/b992d2...
[2]: https://github.com/christianselig/apollo-backend/blob/b992d2...