Deployment types

OpenVidu offers user-friendly installers that facilitate quick on-premises deployments, so you can self-host your real-time solution in your own infrastructure or any cloud provider.

There are different deployment options available, depending on your needs:

Type of deployment	OpenVidu Local (development)	OpenVidu Single Node	OpenVidu Elastic	OpenVidu High Availability
OpenVidu Edition	COMMUNITY PRO	COMMUNITY	PRO	PRO
Suitability	For local development in your laptop	For applications with medium user load	For applications with dynamic user load that require scalability	For applications where both scalability and fault tolerance are critical
Features	Friendly Docker Compose setup with Redis, Egress, Ingress, S3 storage and observability. With automatic certificate management to test across devices in your network	Custom LiveKit distribution with Redis, Egress, Ingress, S3 storage and observability	Same benefits as OpenVidu Single Node plus 2x performance, scalability and advanced observability	Same benefits as OpenVidu Single Node and OpenVidu Elastic plus fault tolerance
Number of servers	Your laptop	1 Node	1 Master Node + N Media Nodes	4 Master Nodes + N Media Nodes
Installation instructions	Install	Install	Install	Install

OpenVidu Local (development)#

To run OpenVidu in your local machine, this is the quickest option. It is a Docker Compose setup that includes all the necessary services to run OpenVidu in your LAN, including automated SSL certificates that will be valid across all devices in your network.

OpenVidu Single Node#

This is the simplest production-ready OpenVidu deployment available. It provides all the features you need, but lacks scalability and fault tolerance. But make no mistake about it: it is perfectly suitable for medium-scale production deployments. For most projects OpenVidu Single Node will be enough, at least until your user load gets serious. You can host hundreds of simultaneous participants in your rooms by running OpenVidu Community on a sufficiently powerful server!

It is composed of a single OpenVidu Node hosting all the necessary services in a monolithic setup.

OpenVidu Elastic#

This is the intermediate OpenVidu deployment. It provides scalability for your video rooms. Suitable for applications with dynamic load in the media plane that require scalability.

It is composed of two different types of nodes, one of them running on a cluster of multiple servers and the other running as a single monolithic server:

A cluster of Media Nodes hosting all the media-related services. Your video rooms scale up and down thanks to this cluster.
A single Master Node hosting all the support services in a monolithic setup.

OpenVidu High Availability#

This is the most complete OpenVidu deployment. It provides scalability for your video rooms and fault tolerance in all its services. Suitable for applications where both scalability and availability are critical.

It is composed of two different types of nodes running on two separate clusters:

A cluster of Media Nodes hosting all the media-related services. Your video rooms scale up and down thanks to this cluster. The minimum number of nodes in this cluster is 1, and it is designed to scale up and down dynamically according to workload.
A cluster of Master Nodes hosting all the support services in their high availability format. Your deployment is fault tolerant thanks to this cluster. The minimum number of nodes in this cluster is 4, and it is designed to have a fixed number of nodes at all times.

Node services#

OpenVidu is composed of several services that work together to provide a complete videoconferencing solution. Every service runs as a Docker container, coordinated with Docker Compose.

Master Node services#

SERVICE	DESCRIPTION
OpenVidu Dashboard	Web application interface for managing your cluster and visualizing your Rooms.
Default Application (OpenVidu Call)	A fully fledged videoconference web application ready to be used out-of-the-box.
OpenVidu Operator	Module that supervises the high availability services and updates the loadbalancing configuration dynamically.
Redis	Database used to share transient information between Media Nodes and coordinate them. In OpenVidu High Availability this is an instance of a Redis Cluster.
MongoDB	Database used to store analytics and monitoring persistent data. In OpenVidu High Availability this is an instance of a MongoDB Replica Set.
Minio	S3 bucket used to store recordings and common node configurations. In OpenVidu High Availability this is an instance of a Minio Multi-Node.
Caddy	Reverse proxy used as a loadbalancer to distribute client connections across your nodes and automatically manage your TLS certificate.
Mimir (observability)	Module used to store metrics from Prometheus.
Promtail (observability)	Module used to collect logs from all services and send them to Loki.
Loki (observability)	Module used to store logs.
Grafana (observability)	Module used to visualize logs and metrics in dashboards.

Media Node services#

SERVICE	DESCRIPTION
OpenVidu Server	Media server used to stream real-time video, audio and data. Based on SFUs LiveKit and mediasoup.
Egress Server	Module used to export media from a Room (for example, recordings or RTMP broadcasting). See Egress.
Ingress Server	Module used to import media into a Room (for example, an MP4 video or an RTSP stream). See Ingress.
Caddy	Reverse proxy used as a loadbalancer to distribute the load generated by the Media Nodes over the Minio, Mimir and Loki cluster.
Prometheus (observability)	Module used to collect metrics from OpenVidu Server and send them to Loki.
Promtail (observability)	Module used to collect logs from all services and send them to Loki.