https://github.com/hystax/optscale-k8s-cost-metrics-collector

https://github.com/hystax/optscale-k8s-cost-metrics-collector

Last synced: about 2 months ago
JSON representation

• Host: GitHub
• URL: https://github.com/hystax/optscale-k8s-cost-metrics-collector
• Owner: hystax
• License: apache-2.0
• Created: 2021-05-11T05:46:38.000Z (over 3 years ago)
• Default Branch: main
• Last Pushed: 2023-01-16T07:44:02.000Z (almost 2 years ago)
• Last Synced: 2024-04-27T00:18:42.364Z (8 months ago)
• Language: Smarty
• Size: 94.7 KB
• Stars: 0
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
#  (obsolete) OptScale Kubernetes Cost Metrics Collector

> This approach of k8s cost metrict collection for OptScale is obsolete due to multiple architectural flaws. Reworked implementation can be found here - https://github.com/hystax/helm-charts/tree/main/charts/kube-cost-metrics-collector

Ready-to-deploy setup of components necessary for kubernetes cost management and FinOps in Hystax OptScale.

Components

- Prometheus

- Prometheus node-exporter module

- kube-state-metrics

- nginx (optional, for setting up authentication)

## Prerequisites

- Kubernetes 1.16+

## Description

Helm chart for the OptScale Kubernetes Collector project, which is created to collect Kubernetes resources information and share it with OptScale FinOps project - https://hystax.com/optscale/.

## Kubernetes YAML (LoadBalancer based configuration)

These resources can be deployed with the following single command:

`kubectl apply -f https://raw.githubusercontent.com/hystax/optscale-k8s-collector-helm-chart/master/optscale-agent.yaml`

## Helm install configurations

### Helm Prerequisites

- Helm 3+

### NodePort example (recommended for virtual machines based clusters):

`helm install --namespace optscale --create-namespace 

--set nginx.service.type=NodePort 

--set nginx.service.enabledExternalPort=true 

--set nginx.service.externalPort=31222 

--set nginx.passwd=$(htpasswd -nbm "optscale" "") 

monitoring .`

### LoadBalancer example (recommended for cloud based clusters):

`helm install --namespace optscale --create-namespace 

--set nginx.service.internalPort=4433 

--set nginx.passwd=$(htpasswd -nbm "optscale" "") 

monitoring .`

## Helm chart parameters

The following table lists the commonly used configurable parameters of the Helm chart and their default values.

Parameter | Description | Default

--------- | ------------------------------------------------ | -------

nginx.enabled | if `false`, use another way to expose the Prometheus port | `true`

nginx.service.type | `NodePort` or `LoadBalancer` | `LoadBalancer`

nginx.service.internalPort | Internal NGINX Server Port Number | `4433`

nginx.passwd | MD5 hash of the username and password to login to the Agent metrics | `optscale:`

nginx.service.enabledExternalPort | use `true` for `NodePort` and `false` - `LoadBalancer`| `false`

nginx.service.externalPort | Agent external Node port number | `31222`