Skip to content

Latest commit

 

History

History
96 lines (71 loc) · 6.72 KB

README.md

File metadata and controls

96 lines (71 loc) · 6.72 KB
Raw

P.S. PRs with fixes or improves are welcome ;-)

About

A Jitsi Meet Chart for Kubernetes with 2 replicas of HAproxy in peering. All Jitsi shard are in backend via server-template in configuration.

This chart will install and configure: HAProxy statefulset, Jicofo deployment, JVB statefulset (scalable), web interface, prosody XMPP server.

Architecture

The Jitsi Kubernetes namespace has the following architecture:

Architecture Jitsi Meet

The setup shown above contains only a single shard (for visual clarity). Subsequent shards would be attached to the web service. A more detailed explanation of the system architecture with multiple shards can be found in architecture/architecture.md.

Install chart

To download and to install the Jitsi Meet Chart, make sure that you have the Helm CLI (v2+) installed and clone this repository on your machine.

To install the chart, in your terminal, go to the jitsi-helm repository and run the following command:

helm install -n jitsi ./ --wait

If you want to use your own custom watermark, first create configmap from your png file:

kubectl create configmap -n <namespace> watermark --from-file=watermark.png

It assumes that you have a Nginx Ingress controller and you use CertManager along with ACME issuer type for managing the HTTPS certificates.

Because --wait flag, the status will be given once Jisti Meet is ready.

Updating the chart

To update the chart, in your terminal, go to the jitsi-helm repository and run the following command:

helm upgrade jitsi ./ --wait

Delete the chart

To delete the chart, in your terminal, go to the jitsi-helm repository and run the following command:

helm delete --purge jitsi

Configuration

The following table lists the configurable parameters of the Jitsi Meet chart and their default values.

Parameter Description Default
shardCount Number of shards 2
haproxy.image Docker image haproxy:2.1
ingress.enabled Enable ingress true
ingress.hosts List of hosts in this ingress empty
ingress.class Which ingressClassName to use empty
ingress.tls.enabled Enable TLS for ingress true
ingress.tls.secretName Name of the secret storing the TLS certificate and key jitsi-tls
ingress.extraPaths Extra paths to add to the ingress []
jicofo.image Jicofo docker image jitsi/jicofo
jicofo.imagePullPolicy Jicofo image pull policy Always
jicofo.extraEnvs Jicofo extra environment variables []
jvb.image JVB docker image jitsi/jvb
jvb.image.imagePullPolicy JVB image pull policy Always
jvb.replicas JVB replica count 1
jvb.monitoringEnable JVB exporter container true
jvb.hostPort JVB hostPort empty
jvb.nodeportPrefix JVB Node port prefix 30
jvb.extraEnvs JVB extra environment variables []
prosody.image Prosody docker image jitsi/prosody
prosody.image.imagePullPolicy Prosody image pull policy Always
prosody.extraEnvs Extra env var for prosody deployment []
prosody.extraVolumes Additionnal volumes to the prosody deployment []
prosody.extraVolumeMounts Additional volume mounts to the prosody deployment []
prosody.globalModules Additional global modules to enable on prosody []
prosody.globalConfig Additional global config parameters on prosody []
uvs.enable Whether the Matrix User Verification Service should be enabled false
web.image Web docker image jitsi/web
web.image.imagePullPolicy Web image pull policy Always
web.extraEnvs Extra env var for web deployment []
web.extraVolumes Additionnal volumes to the web deployment []
web.extraVolumeMounts Additional volume mounts to the web deployment []
watermark Watermark logo true

Running two jitsi instances inside the same cluster

  1. The second instance should be deployed with specific settings :
  • jvb.nodeportPrefix should use a different value from 30 to avoid ports conflicts or
  • jvb.hostPort that's distinct from other installations