Enable Serverless Metrics in Apache OpenWhisk on Kubernetes with Prometheus
Serverless functions are event-triggered, stateless and ephemeral, which makes metrics essential to a Serverless platform. Both system metrics and user metrics are helpful for operators and developers …
| Talk Title | Enable Serverless Metrics in Apache OpenWhisk on Kubernetes with Prometheus |
| Speakers | Ying Chun Guo (Software Engineer, IBM) |
| Conference | KubeCon + CloudNativeCon |
| Conf Tag | |
| Location | Shanghai, China |
| Date | Jun 23-26, 2019 |
| URL | Talk Page |
| Slides | Talk Slides |
| Video | |
Serverless functions are event-triggered, stateless and ephemeral, which makes metrics essential to a Serverless platform. Both system metrics and user metrics are helpful for operators and developers to understand what happens in the server side. OpenWhisk is an open sourced Serverless/Function-as-a-Service (FaaS) platform that can be deployed on Kubernetes. Apache OpenWhisk uses Kamon to collect metrics by default. Kamon Prometheus exporter supports for exporting the metrics to Prometheus, which is the popular Kubernetes monitoring tool. This presentation will introduce what kind of metrics are defined on Apache OpenWhisk, then show how to enable these metrics on Kubernetes using Prometheus and Grafana, and at last demonstrate a live demo.
