Problem

Checkly currently runs your checks in a round-robin fashion, cycling through the locations you have selected for your checks. This is an "easy to reason about" system.

However, for global infrastructure and apps like CDNs, Anycast-type systems, or any app that resolves one domain to region specific endpoint this is not great as a failing check that fails in one specific region, will resolve if the next region is healthy. The result is that the check does not indicate of one specific region is having problems, whereas others are fine.

Proposed Solution

• Checkly should have the option to schedule checks parallel for all regions, instead of round-robin.

• The status of a check should be able to have a "partial" or "regional" outage.

• Checkly should be able to alert on regional outages

  original GH issue link https://github.com/checkly/public-roadmap/issues/198