HazardB / is-kronos-up-yet

Website for the "is kronos up" service
MIT License
6 stars 0 forks source link

Enhance the API and worker to provide latency stats #12

Open HazardB opened 8 years ago

HazardB commented 8 years ago

Average response time would be a great new feature, especially if a small graph can be provided showing the last 5 minutes.

This would require increasing the timeout on the TCP connection - we could probably go to 10 seconds.

We would need to change the UI to display any sever with a response time over 2s as "orange" aka "unstable". The brackets could work as follows:

Up: 0-2000ms Unstable: 2k-6k ms Down: >6k ms

Some database optimisation would be needed, as well as API changes to support retrieving historic status records.