tnolet
commented
2 years ago @IanWhitney thanks for contributing. This smells a lot like "bucket based" / "budget based" alerting or SLOs. The immediate question I would have is "p99 over how much time?" as it is an aggregate metrics over time. Quickly you will end up with either a sliding window or a fixed error budget per calendar time.
It is similar enough to this ticket that I'm referencing it here https://github.com/checkly/public-roadmap/issues/186
💡 For general support requests and bug reports, please go to checklyhq.com/support
Is your feature request related to a problem? Please describe. Instead of alerting on a single request being slow, I'd like to alert if the average or p95 time exceeds a limit.
Describe the solution you'd like When setting Assertions, I would like to be able to chose "average response time" or "p95 response time" as the Property value for a Response Time check.
Describe alternatives you've considered Checking the dashboard manually? I'm not sure what other options there are.
Additional context A single slow request may not be worth an alert. But several slow requests will raise our average/p95 response times and are worth investigating.