Skip to content

Latest commit

 

History

History
24 lines (15 loc) · 1.7 KB

README.md

File metadata and controls

24 lines (15 loc) · 1.7 KB
Raw

Vector Helm Chart

DEPRECATED This chart has been deprecated for the vector chart.

This is an opinionated Helm Chart for running Vector as an aggregator in Kubernetes.

To get started check out our documentation

Load Balancing (Beta)

As of the 0.15.0 release we've included a HAProxy deployment to load balance traffic to Vector instances running in the aggregator role. Using an external proxy allows us to scale aggregators horizontally in a source agnostic way, making it easy to migrate existing infrastructure to use Vector.

To enable the HAProxy deployment you can pass --set haproxy.enabled=true with your helm command1, by default this will create frontend and server-template configurations generated from your service.ports entries. Assuming a release name of "vector" clients sending events to an aggregator with HAProxy enabled should the hostname of "vector-aggregator-haproxy". With HAProxy we leverage its builtin resolvers configuration and Kubernetes DNS to dynamically populate the backend Vector instances.

Our default configurations are designed to work out of the box, but we recommend production deployments should optimize their configuration based on the specific needs of that environment and configuration.

haproxy:
  enabled: true

Footnotes

  1. Or set it in your values file with: ↩