BharatSahAIyak / k8s

0 stars 0 forks source link

map kong metrics to prometheus and view those in grafana #95

Open Srijan-SS02 opened 1 month ago

Srijan-SS02 commented 1 month ago

prepare/copy official dashboard

meeting with @singhalkarun and @GJS2162 (for 30min)

Srijan-SS02 commented 3 weeks ago

Updates

max RPS

old: max(kong_http_requests_total{service=~"$service", route=~"$route"})
new: max(rate(kong_http_requests_total{service=~"$service", route=~"$route"}[$__interval]))
reason: kong_http_requests_total metric is a cumulative counter of HTTP requests, which means it increases monotonically over time. If you're looking for requests per second (RPS), you typically want to calculate the rate of increase over a time window. This can be done using the rate() function. Grafana Prometheus

avg RPS

same changes as max RPS, rate is added.

min RPS

same changes as max RPS, rate is added.

For Yxx

dynamic interval is added [$__interval], earlier it was fixed to static [1m]
this enables updation of values, when time period is changed
3xx error percentage fixed

For latency

dynamic interval is added [$__interval], earlier it was fixed to static [1m]
this enables updation of values, when time period is changed

Srijan-SS02 commented 3 weeks ago

Srijan-SS02 commented 3 weeks ago

PR: https://github.com/Samagra-Development/devops/pull/86