Skip to content
This repository has been archived by the owner on Nov 8, 2022. It is now read-only.

Demo flow

Jump to bottom
Andrzej Kuriata edited this page Apr 19, 2016 · 5 revisions

For discussions regarding demo flow please visit: https://github.com/intelsdi-x/kubesnap/issues/1

  1. Start a Kubernetes cluster
  2. Start PoC-snap-Pod on a minion
  3. Start InfluxDB-and-Grafana Pod to show that snap metrics are retrieved by Heapster
  4. Run PoC-Heapster which is able to consume data exposed by snap (snap as the only data source)
  5. Start PoC-Pod which will be under HPA supervision
  6. PoC-Pod is starting to utilize more CPU (for example: using http://linux.die.net/man/1/stress)
  7. CPU utilization graph in Grafana for PoC-Pod is showing increased CPU utilization by PoC-Pod and by node in general
  8. HPA spawns new PoC-Pods

(Possible continuation)

  1. Load new snap collector plugin
  2. Show that new graphs, for new metrics, can be added in Grafana
  3. Modify HPA to support other scaling criteria (new parameter similar --cpu-percent) to showcase use of new metrics
Clone this wiki locally