hypothesis / hypothesis/product-backlog
Outline a way to remove "badge misses" from h
- Dominant language
- No language data
- Stars
- 122
- Forks
- 7
- PR merge metrics
- No merged PRs in 30d
Description
This is a spike that came out of this [earlier spike about the badge](https://github.com/hypothesis/product-backlog/issues/953).
**Problem 1:** We don't think that "badge misses" (those badge requests where the badge number is 0 because the URL in question hasn't been annotated) belong in h because this traffic has very different scaling properties than the rest of h. It scales with the number of Chrome extension installs and the amount of web browsing done by those users, regardless of whether those users are using our extension or are even logged in. It doesn't scale along with actual usage of h. And this is different to how all the rest of h's traffic scales.
We think it's okay to remove the badge misses only and let badge hits (when the badge number is or might be greater than 0) to still go to h. Badge hits likely scale more similarly to the rest of h's traffic and are also just a lot less traffic.
**Problem 2:** All the badge misses create a lot of noise on all of h's logging and monitoring systems. For example in Sentry and New Relic.
**The task for this spike** is to investigate and outline a solution to these two problems and type it up on this card.
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.