Deploying OpenVidu Enterprise on premises


Single master deployment 🔗


OpenVidu Enterprise on premises with a single master is deployed just as OpenVidu Pro on premises. You just need to add the following property to OpenVidu configuration .env file here.

OPENVIDU_EDITION=enterprise


High Availability deployment 🔗


Overview 🔗

OpenVidu Enterprise on premises can be deployed in a High Availability (HA) configuration. The difference of OpenVidu Enterprise Single Master deployment and OpenVidu Enterprise HA deployment is that in the latter you will have several OpenVidu Enterprise Nodes running at the same time, and all of them will be able to handle your users' sessions. This is specially useful when you have a lot of sessions and you want to distribute the load among several servers.

The following diagram shows a typical HA deployment:

You need a Load Balancer (LB) to distribute the traffic among the different OpenVidu Enterprise Nodes. The LB will be the only entry point to your OpenVidu Enterprise deployment. The LB will redirect the traffic to the different OpenVidu Enterprise Nodes, and each of them will be able to handle the signaling plane. The LB will also be in charge of detecting when a Node is down, and redirecting the traffic to the rest of the Nodes.

The media plane will be handled directly by the OpenVidu Enterprise Nodes. The LB will not be involved in the media traffic.

To be fully Highly Available, you need to deploy the following services also HA:

  • Redis (for storing sessions and cluster information)
  • S3 (for storing recordings and configuration)
  • Elasticsearch and Kibana (for storing and visualizing logs and metrics)

To simplify the deployment explanation, we will explain a reduced version of the HA deployment, where all the services, Load Balancer and OpenVidu Call will be deployed in a node called Base Services, and the OpenVidu Enterprise Nodes will be deployed in different nodes. But keep in mind that the Base Services node is not production ready, and it is only used here to simplify the explanation.

This is the diagram of the reduced HA deployment:

Prerequisites 🔗

  • You must have at least 3 different machine/instances with a Public IP address each one. One of them will be the Base Services node, and the rest will be the OpenVidu Enterprise Nodes.
  • Each machine must have:
    • Any Linux distribution (we recommend Ubuntu)
    • Docker and Docker Compose installed
    • A public IP address.
    • All the instances should be in the same network, and they should be able to communicate with each other (Ports configuration will be explained later).
    • Each instance must have at least 8GB of RAM and 4 CPU cores. We recommend having 16GB of RAM and 8 CPU cores and some extra disk space for recordings. Recordings are stored locally while being recorded, and then they are uploaded to S3. So you need to have enough disk space to store all the recordings while they are being recorded.
  • A FQDN (Fully Qualified Domain Name) pointing to the Base Services node. This FQDN will be used to access OpenVidu Call and as a Load Balancer for OpenVidu Enterprise.

In the following image you can see the private IPs addresses of each node. We will use this network setup in the rest of the deployment explanation:

Base Services ports configuration 🔗

The following are the inbound and outbound rules that you need to open in the Base Services node:


Inbound rules 🔗
  • (Optional) Port 22 TCP (Open to 0.0.0.0,::/0): Used to access the machine via SSH for administration purposes.
  • Port 80 TCP (Open to 0.0.0.0,::/0): Used by the nginx proxy for letsencrypt certificate generation.
  • Port 443 TCP (Open to 0.0.0.0,::/0): Used by the nginx proxy to serve OpenVidu Call and OpenVidu Enterprise API.
  • Port 6379 TCP (Open to 10.5.0.0/24): Used by Redis to store sessions and cluster information. This port should be open only to the OpenVidu Enterprise Nodes.
  • Port 5601 TCP (Open to 10.5.0.0/24): Used by Kibana to serve the logs and metrics. This port should be open only to the OpenVidu Enterprise Nodes.
  • Port 9000 TCP (Open to 10.5.0.0/24): Used by Minio for configuration and recordings. This port should be open only to the OpenVidu Enterprise Nodes.
  • Port 9200 TCP (Open to 10.5.0.0/24): Used by Elasticsearch to serve the logs and metrics. This port should be open only to the OpenVidu Enterprise Nodes.
  • Port 5443 TCP (Open to 10.5.0.0/24): Used to serve the OpenVidu Enterprise API through internal network. If you have your OpenVidu Application Server in the same network, you can use http://10.5.0.5:5443 to access the API, instead of using the external domain name. This is only valid for OpenVidu Application Servers, not for browser/device clients.


Outbound rules 🔗

OpenVidu Enterprise Nodes ports configuration 🔗

The following are the inbound and outbound rules that you need to open in the OpenVidu Enterprise Nodes:


Inbound rules 🔗
  • (Optional) Port 22 TCP (Open to 0.0.0.0,::/0): Used to access the machine via SSH for administration purposes.
  • Port 443 TCP/UDP (Open to 0.0.0.0,::/0): This port in OpenVidu Enterprise HA is used for TURN/STUN instead of 3478.
  • Port 3000 TCP: (Open to 10.5.0.0/24): Used by a service called Media Node Controller which provision the Media Servers and recording containers. This port should be open only to the OpenVidu Enterprise Nodes.
  • Port 4000 TCP: (Open to 10.5.0.0/24): Used by the Speech To Text service. This port should be open only to the OpenVidu Enterprise Nodes.
  • Port 4443 TCP: (Open to 10.5.0.0/24): Used by the Replication Manager service to replicate the sessions and recordings and proxy accordingly all the requests to the OpenVidu Server. This port should be open only to the OpenVidu Enterprise Nodes and the Load Balancer.
  • Port 5443 TCP: (Open to 10.5.0.0/24): Used by OpenVidu Server. This port is consumed by the Replication Manager service, and it should be open only to the OpenVidu Enterprise Nodes.
  • Port 40000-65535 TCP/UDP (Open to 0.0.0.0,::/0): Used by Media Server to send and receive media streams.


Outbound rules 🔗

Installing Base Services 🔗

In order to install the base services, you need to follow the next steps:

1) SSH into the Base Services machine (In the example is 10.5.0.5)

2) Change to superuser:

sudo su

3) Go to /opt directory:

cd /opt

4) Now execute the following command to download and run the installation script.

curl https://s3-eu-west-1.amazonaws.com/aws.openvidu.io/install_ov_enterprise_ha_base_latest.sh | bash

This will download all required files into ov-enterprise-base-services folder and will show this message with basic instructions:

=======================================
OpenVidu Enterprise HA successfully installed.
=======================================

1. Go to openvidu folder:
------------------------------------------------
$ cd ov-enterprise-base-services
------------------------------------------------


2. Configure the .env file with your own values:
   Check the documentation for more information:
   https://docs.openvidu.io/en/<version>/deployment/enterprise/on-premises/#high-availability-deployment
------------------------------------------------
$ nano .env
------------------------------------------------


3. Start OpenVidu Enterprise HA Base Services:
------------------------------------------------
$ ./base-services start
------------------------------------------------

To deploy a fixed version, including previous ones, replace latest with the desired version number like:
curl https://s3-eu-west-1.amazonaws.com/aws.openvidu.io/install_ov_enterprise_ha_base_2.29.0.sh | bash

5) Go to ov-enterprise-base-services folder:

cd ov-enterprise-base-services

6) In this directory you will see a .env file. This file contains all the configuration parameters that you need to set up in order to run the Base Services. As this base service contains the Load Balancer, S3, Redis, Elasticsearch and OpenVidu Call Application, we will configure those services via this config file:

  • DOMAIN_OR_PUBLIC_IP: Set the FQDN or public IP of the machine where you are installing the Base Services.
  • OPENVIDU_SECRET: Set the secret that you want to use for the OpenVidu Server. This parameter is needed by the default OpenVidu Call Application to connect to the OpenVidu Server.
  • CERTIFICATE_TYPE: Set the type of certificate that you are using, which can be selfsigned, owncert or letsencrypt. If you want to know more about this parameter, please check this section.
  • LETSENCRYPT_EMAIL: Set the email that you want to use for the certificate.
  • OPENVIDU_ENTERPRISE_HA_NODE_IPS: Set the IP addresses of the OpenVidu Enterprise Nodes, separated by commas.
  • OPENVIDU_ENTERPRISE_HA_REDIS_PASSWORD: Set the password for the deployed Redis database in this machine. This password will be used by the OpenVidu Enterprise Nodes to connect to this Redis database.
  • OPENVIDU_ENTERPRISE_HA_S3_CONFIG_BUCKET: Set the name of the S3 bucket that you want to use to store the configuration files (and recordings optionally). This bucket will be created automatically in a Minio server that will be deployed in this machine and will be used by the OpenVidu Enterprise Nodes.
  • OPENVIDU_ENTERPRISE_HA_S3_CONFIG_ACCESS_KEY: Set the access key for the S3 Minio Service.
  • OPENVIDU_ENTERPRISE_HA_S3_CONFIG_SECRET_KEY: Set the secret key for the S3 Minio Service.
  • ELASTICSEARCH_USERNAME: Set the username for the Elasticsearch and Kibana service.
  • ELASTICSEARCH_PASSWORD: Set the password for the Elasticsearch and Kibana service.

The configuration file for this deployment would look like this:

DOMAIN_OR_PUBLIC_IP=openvidu.example.io
OPENVIDU_SECRET=OPENVIDU_SECRET
CERTIFICATE_TYPE=letsencrypt
LETSENCRYPT_EMAIL=openvidu@example.io
OPENVIDU_ENTERPRISE_HA_NODE_IPS=10.5.0.6,10.5.0.7
OPENVIDU_ENTERPRISE_HA_REDIS_PASSWORD=REDIS_SECRET
OPENVIDU_ENTERPRISE_HA_S3_CONFIG_BUCKET=openvidu-enterprise
OPENVIDU_ENTERPRISE_HA_S3_CONFIG_ACCESS_KEY=minioadmin
OPENVIDU_ENTERPRISE_HA_S3_CONFIG_SECRET_KEY=MINIO_SECRET
ELASTICSEARCH_USERNAME=elasticadmin
ELASTICSEARCH_PASSWORD=ELASTIC_SECRET

Note the following:

  • The domain openvidu.example.io should be pointing to the public IP of the machine where you are installing the Base Services. If you are using a self-signed certificate, you can use the public IP of the machine instead of the domain. If you are using a Let's Encrypt certificate, you need to set the domain.
  • The OPENVIDU_ENTERPRISE_HA_NODE_IPS parameter should contain the IP addresses of the OpenVidu Enterprise Nodes that you are going to deploy. In this example, the IPs are 10.5.0.6 and 10.5.0.7
  • Make sure to change secret values OPENVIDU_SECRET, REDIS_SECRET, MINIO_SECRET and ELASTIC_SECRET to your own values. Use alphanumeric characters, dashes and underscores only, and do not use special characters.

7) Now, you can start the Base Services by executing the following command:

./base-services start

8) After deploying OpenVidu Enterprise HA Base Services, and if the installation of OpenVidu Enterprise Nodes has been successful, you will be able to access the following services:

  • OpenVidu Call Application: https://openvidu.example.io
  • Kibana: https://openvidu.example.io/kibana
  • Minio: https://openvidu.example.io/minio
  • OpenVidu Inspector: https://openvidu.example.io/inspector/

Installing OpenVidu Enterprise Nodes 🔗

This guide provides step-by-step instructions on how to install and set up OpenVidu Enterprise Nodes. We will set up one node first (10.5.0.6), then replicate this process for the second node (10.5.0.7).

1) SSH into the Base Services machine (In the example is 10.5.0.6):

2) Change to superuser:

sudo su

3) Go to /opt directory:

cd /opt

4) Now execute the following command to download and run the installation script:

curl https://s3-eu-west-1.amazonaws.com/aws.openvidu.io/install_openvidu_enterprise_ha_node_latest.sh | bash

To deploy a fixed version, including previous ones, replace latest with the desired version number like:
curl https://s3-eu-west-1.amazonaws.com/aws.openvidu.io/install_openvidu_enterprise_ha_node_2.29.0.sh | bash

This will download all required files into openvidu folder and will show this message with basic instructions:

=======================================
OpenVidu Enterprise HA successfully installed.
=======================================

1. Go to openvidu folder:
------------------------------------------------
$ cd openvidu
------------------------------------------------


2. Configure the .env file with your own values:
   Check the documentation for more information:
   https://docs.openvidu.io/en/<version>/deployment/enterprise/on-premises/#high-availability-deployment
------------------------------------------------
$ nano .env
------------------------------------------------


3. Start OpenVidu
------------------------------------------------
$ ./openvidu start
------------------------------------------------

5) Go to openvidu folder:

cd openvidu

6) In this directory, you will find the .env file. This file contains all the configuration parameters that you need to set for the OpenVidu Enterprise Node. The minimal configuration that you need to set is the following:

  • DOMAIN_OR_PUBLIC_IP: Set the domain or public IP of the machine where you are installing the OpenVidu Enterprise Node. This domain name should be the same as the one used by the Load Balancer.
  • OPENVIDU_PRO_LICENSE: Set the license of your OpenVidu License subscription. You need an OpenVidu account to purchase it. There's a 15 day free trial waiting for you!
  • OPENVIDU_SECRET: The OpenVidu Secret used by Applications to connect to the OpenVidu Enterprise Node.
  • OPENVIDU_ENTERPRISE_MEDIA_SERVER: The media server used by OpenVidu Enterprise Node. You can choose between kurento or mediasoup.
  • HTTPS_PORT: The port used by the Load Balancer to redirect the traffic to the OpenVidu Enterprise Node. This port should be the same as the one used in the Load Balancer.
  • OPENVIDU_ENTERPRISE_HA_NODE_PRIVATE_IP: The private IP of the machine where you are installing the OpenVidu Enterprise Node.
  • OPENVIDU_ENTERPRISE_HA_REDIS_HOST: The IP or hostname of your Redis deployment.
  • OPENVIDU_ENTERPRISE_HA_REDIS_PASSWORD: The password of your Redis deployment.
  • OPENVIDU_ENTERPRISE_HA_S3_CONFIG_SERVICE_ENDPOINT: The endpoint of your S3 deployment/provider.
  • OPENVIDU_ENTERPRISE_HA_S3_CONFIG_BUCKET: The bucket name you want to use to store the configuration files.
  • OPENVIDU_ENTERPRISE_HA_S3_CONFIG_ACCESS_KEY: The access key of your S3 deployment/provider.
  • OPENVIDU_ENTERPRISE_HA_S3_CONFIG_SECRET_KEY: The secret key of your S3 deployment/provider.
  • OPENVIDU_ENTERPRISE_HA_S3_CONFIG_REGION: The region of your S3 provider. (Usually with on premises deployments like Minio, this value is not required).
  • OPENVIDU_PRO_ELASTICSEARCH_HOST: The http URL of your Elasticsearch deployment/service.
  • OPENVIDU_PRO_KIBANA_HOST: The http URL of your Kibana deployment/service.
  • ELASTICSEARCH_USERNAME: The username of your Elasticsearch deployment/service.
  • ELASTICSEARCH_PASSWORD: The password of your Elasticsearch deployment/service.

The configuration file for this deployment would look like this:

DOMAIN_OR_PUBLIC_IP=openvidu.example.io
OPENVIDU_PRO_LICENSE=<YOUR_OPENVIDU_PRO_LICENSE>
OPENVIDU_SECRET=OPENVIDU_SECRET
OPENVIDU_ENTERPRISE_MEDIA_SERVER=mediasoup
HTTPS_PORT=443
OPENVIDU_ENTERPRISE_HA_NODE_PRIVATE_IP=10.5.0.6
OPENVIDU_ENTERPRISE_HA_REDIS_HOST=10.5.0.5
OPENVIDU_ENTERPRISE_HA_REDIS_PORT=6379
OPENVIDU_ENTERPRISE_HA_REDIS_PASSWORD=REDIS_SECRET
OPENVIDU_ENTERPRISE_HA_S3_CONFIG_SERVICE_ENDPOINT=http://10.5.0.5:9000
OPENVIDU_ENTERPRISE_HA_S3_CONFIG_BUCKET=openvidu-enterprise
OPENVIDU_ENTERPRISE_HA_S3_CONFIG_ACCESS_KEY=minioadmin
OPENVIDU_ENTERPRISE_HA_S3_CONFIG_SECRET_KEY=MINIO_SECRET
OPENVIDU_PRO_ELASTICSEARCH_HOST=http://10.5.0.5:9200
OPENVIDU_PRO_KIBANA_HOST=http://10.5.0.5:5601
ELASTICSEARCH_USERNAME=elasticadmin
ELASTICSEARCH_PASSWORD=ELASTIC_SECRET

Observe that the OPENVIDU_ENTERPRISE_HA_NODE_PRIVATE_IP is the private IP of the machine where you are installing the OpenVidu Enterprise Node, and the rest of the configuration parameters for Redis, S3, Elasticsearch and Kibana are pointing to the private IP of the machine where you had installed the Base Services.

7) Start the OpenVidu Enterprise Node:

./openvidu start

When you see the following message, the OpenVidu Enterprise Node is ready to be used:

replication-manager  | ----------------------------------------------------
replication-manager  |
replication-manager  |    OpenVidu Enterprise HA Node is ready!
replication-manager  |    ---------------------------
replication-manager  |
replication-manager  |    * OpenVidu Server URL: https://openvidu.example.io:443/
replication-manager  |
replication-manager  |    * OpenVidu Dashboard: https://openvidu.example.io:443/inspector
replication-manager  |
replication-manager  |    NOTE: This is a node of an OpenVidu Enterprise Cluster
replication-manager  |      with High Availability. Your Load Balancer configured
replication-manager  |      with the domain name 'openvidu.example.io'
replication-manager  |      should register the private IP of this node: '10.5.0.6'
replication-manager  |
replication-manager  | ----------------------------------------------------

8) To set up additional nodes, simply replicate the installation process previously detailed. Copy the .env file to the new machines, initiate the OpenVidu Enterprise Node, and modify the OPENVIDU_ENTERPRISE_HA_NODE_PRIVATE_IP to match the private IP of the current installation machine. In our example we would have to set the following value in the .env file of the second node:

OPENVIDU_ENTERPRISE_HA_NODE_PRIVATE_IP=10.5.0.7

Node Management 🔗

To effectively manage the scalability of your cluster, it is crucial to monitor the CPU and RAM usage of your OpenVidu Enterprise Nodes and scale up or down depending on those. This section provides a comprehensive guide on how to perform necessary actions for scaling your cluster up or down, but the monitoring strategy is up to you.

Scale Up 🔗

Follow the steps below to scale up your system:

1) Create a New Machine:

First, create a new machine and ensure it has a public IP. This machine will be used to install the OpenVidu Enterprise Node.

2) Install the OpenVidu Enterprise Node:

Install the OpenVidu Enterprise Node as stated in the Installation section.

Locate the .env file on the new machine and set the OPENVIDU_ENTERPRISE_HA_NODE_PRIVATE_IP parameter. This should be the private IP of the machine you are currently installing. For example, if the private IP of the new machine is 10.5.0.8, your .env file should contain:

OPENVIDU_ENTERPRISE_HA_NODE_PRIVATE_IP=10.5.0.8

3) Update your Load Balancer:

Update your Load Balancer configuration by adding the private IP of the new machine.

If you are using the Base Services as a Load Balancer, find the .env file on the Base Services machine. Modify the OPENVIDU_ENTERPRISE_HA_NODE_IPS parameter by adding the private IP of the new machine you've just set up. For example, if you're adding a new machine with the private IP 10.5.0.8 and your current IPs are 10.5.0.6 and 10.5.0.7, the parameter should look like this:

OPENVIDU_ENTERPRISE_HA_NODE_IPS=10.5.0.6,10.5.0.7,10.5.0.8

Apply the changes. By executing the following command you will update the Load Balancer without interrupting current client connections:

./base-services update-loadbalancer

You've now successfully scaled up your system by adding a new machine to the load balancer. To check if your Node was registered, The message of the log of the OpenVidu Enterprise Node should show this:

OpenVidu Enterprise HA Node is ready!

To receive and event when the Node is registered into the cluster, enable webhooks in the configuration and enable the event HANodeRegistered. You will receive an HTTP POST request like this when the Node is registered:

{
  "timestamp": 1687190122864,
  "event": "HANodeRegistered",
  "nodeId": "master_10.5.0.8",
  "ip": "10.5.0.8"
}

Note: Using any Load Balancer from any Cloud provider, you can check if a node is available by asking directly to the nodes to this specific path: http://<NODE_IP>:4443 or http://<NODE_IP>:4443/openvidu/health. If the node is available, it will return a 200 status code. If not, basically no response will be received.

Scale Down 🔗

The process of scaling down typically presents a greater complexity compared to scaling up due to potential active sessions within nodes. It is vital to maintain the continuity of these sessions, ensuring there are no interruptions. However, OpenVidu Enterprise HA offers a mechanism that addresses this challenge by waiting for ongoing sessions to conclude prior to node shutdown, ensuring a secure and safe termination of the node.

In order to accomplish this, the following steps should be adhered to:

1) Put the node in a waiting-idle-to-terminate state:

Identify the node you want to remove from the cluster and put it in a waiting-idle-to-terminate state. This can be done by executing an HTTP PATCH request to /openvidu/api/media-nodes/<MEDIA_NODE_ID> endpoint of the Load Balancer (see REST API for more information). The <MEDIA_NODE_ID> is the ID of the node you want to remove from the cluster and the body of the request should contain the following JSON:

{
  "status": "waiting-idle-to-terminate"
}

This will put the node in a state which will not accept new sessions, but will wait for ongoing sessions to conclude before terminating. When the node has no active sessions, it will automatically terminate, and terminate all related processes, so the node is no longer part of the cluster and can be safely removed from your infrastructure.

Once in this state, you can't change the status of the node to any other state. If you want to abort the termination process, you need to wait for the node to terminate and then start it again with:

./openvidu restart

To receive an event when the Node is terminated, enable webhooks in the configuration and enable the event HANodeTerminated. You will receive an HTTP POST request like this when the Node is terminated:

{
  "timestamp": 1687190122864,
  "event": "HANodeDeregistered",
  "nodeId": "master_10.5.0.8",
  "ip": "10.5.0.8"
}

2) Update your Load Balancer:

Update your Load Balancer configuration by removing the private IP of the node you want to remove from the cluster. Depending on your Load Balancer, it should be detected that the node is no longer available, and it should be removed from the cluster.

If you are using Base Services as a Load Balancer, find the .env file on the Base Services machine. Modify the OPENVIDU_ENTERPRISE_HA_NODE_IPS parameter by removing the private IP of the node you want to remove from the cluster. For example, if you're removing the node with the private IP 10.5.0.8 the configuration should look like this:

OPENVIDU_ENTERPRISE_HA_NODE_IPS=10.5.0.6,10.5.0.7

Apply the changes. By executing the following command you will update the Load Balancer without interrupting current client connections:

./base-services update-loadbalancer

You've now successfully scaled up your system by adding a new machine to the load balancer.

Note: Using any Load Balancer from any Cloud provider, you can check if a node is available by asking directly to the nodes to this specific path: http://<NODE_IP>:4443 or http://<NODE_IP>:4443/openvidu/health. If the node is available, it will return a 200 status code. If not, basically no response will be received.

3) Remove the node from your infrastructure:

Once the node has been terminated, you can safely remove it from your infrastructure without any issues.

Configuration management 🔗

OpenVidu Enterprise HA provides a mechanism to manage the configuration of the cluster. This mechanism is managed using the deployed and configured S3 bucket.

When you execute ./openvidu start, a process of sync with S3 is runned next to a mechanism to check Redis and S3 are reachable. When you start your first node, no configuration is stored yet in the S3 bucket yet. You will notice that because you will see a message like this in the logs:

    => Configuration file not found in S3

Until the S3 bucket is not reachable, the configuration is managed at /opt/openvidu/.env. Once the S3 bucket is reachable, the configuration is managed at the S3 bucket or via REST API, except the parameter OPENVIDU_ENTERPRISE_HA_NODE_PRIVATE_IP that is managed at /opt/openvidu/.env always. When your configuration is in sync with the node, you will see this message in the logs:

    => Configuration file found in S3
    => Configuration file updated from S3 to local file /opt/openvidu/.env

You can update the configuration through either the REST API or the S3 bucket. If you opt to use the REST API, simply send an HTTP POST request to /openvidu/api/restart along with the updated configuration in the request body. This approach automatically updates the S3 bucket and triggers a restart for all nodes, ensuring the application of the changes.

However, if you prefer to alter the configuration directly via the S3 bucket, you will need to manually restart the nodes. You can do this by executing ./openvidu restart or by sending an HTTP POST request to /openvidu/api/restart. In this latter case, though, remember not to include the configuration in the request body.

In the following sections, we will give two examples of configuration updates via HTTP POST requests to enable recordings and to enable webhooks.

Enabling Recording 🔗

By default, recordings are disabled in this OpenVidu Enterprise HA deployment. To enable them, you need to update the configuration. In this example we will use the same S3 bucket we are using to store the configuration, but you can use any other S3 provider and bucket you wish. The HTTP POST request to /openvidu/api/restart should look like this:

{
    "OPENVIDU_RECORDING": "true",
    "OPENVIDU_PRO_AWS_S3_BUCKET": "openvidu-enterprise",
    "OPENVIDU_PRO_AWS_S3_SERVICE_ENDPOINT": "http://10.5.0.5:9000",
    "OPENVIDU_PRO_AWS_ACCESS_KEY": "minioadmin",
    "OPENVIDU_PRO_AWS_SECRET_KEY": "MINIO_SECRET"
}

Make sure to change the values of the parameters OPENVIDU_PRO_AWS_S3_BUCKET, OPENVIDU_PRO_AWS_S3_SERVICE_ENDPOINT, OPENVIDU_PRO_AWS_ACCESS_KEY and OPENVIDU_PRO_AWS_SECRET_KEY to match your S3 bucket configuration. In this case, as we are using the S3 bucket deployed at Base Services it would look like the previous example.

Enabling Webhooks 🔗

If you want to receive events from OpenVidu Enterprise HA, you need to enable webhooks. To do so, you just need to execute an HTTP POST request to /openvidu/api/restart with the following configuration:

{
  "OPENVIDU_WEBHOOK": "true",
  "OPENVIDU_WEBHOOK_ENDPOINT": "https://<YOUR_WEBHOOK_ENDPOINT>"
}

Make sure to change the value of the parameter OPENVIDU_WEBHOOK_ENDPOINT to match your webhook endpoint.

Troubleshooting 🔗

During the installation process, it is usual sometimes to find some errors. Most of the errors will pop up while installing the system. Some errors you may find are:


Error connecting to S3 bucket 🔗

=> OpenVidu was unable to connect to S3 bucket <bucket-name> to sync the configuration file
    Make sure the configuration of your S3 bucket is correct and that the credentials provided in /opt/openvidu/.env are valid
    Also make sure your S3 bucket is reachable from this machine
    To fix this, follow the next
        1) Return to shell pressing Ctrl+C
        2) Check your configuration file at /opt/openvidu/.env and check the S3 configuration
            NOTE: All S3 parameters starts with 'OPENVIDU_ENTERPRISE_HA_S3_'
        3) Restart OpenVidu with:
            $ ./openvidu

This error will happen if the S3 bucket is not reachable from the machine. Make sure all the OPENVIDU_ENTERPRISE_HA_S3_ parameters are correct and that the S3 bucket is reachable from the machine so the Node can sync the configuration file.


Error connecting to Redis 🔗

 =>  OpenVidu was unable to connect to Redis server <REDIS_HOST>:<REDIS_PORT>

    Make sure the configuration of your Redis server is correct and that the credentials provided in /opt/openvidu/.env are valid
    Also make sure your Redis server is reachable from this machine
    To fix this, follow the next steps:

        1) Return to shell pressing Ctrl+C
        2) Check the file of this node at /opt/openvidu/.env and check the Redis configuration
            NOTE: All redis parameters starts with 'OPENVIDU_ENTERPRISE_HA_REDIS_'
        3) Restart OpenVidu with:
            $ ./openvidu restart

This error will happen if the Redis server is not reachable from the machine. Make sure all the OPENVIDU_ENTERPRISE_HA_REDIS_ parameters are correct and that the Redis server is reachable from the machine.


Other problems or errors 🔗

If none of this errors are happening to you:

1) Check the logs of all the services running in the machine. You can check the logs of the services with the following commands:

sudo su
cd /opt/openvidu
./openvidu logs

Or check the logs of a specific service, for example of replication-manager and openvidu-server

sudo su
cd /opt/openvidu
docker-compose logs openvidu-server
docker-compose logs replication-manager

2) Make sure all your nodes has a reachable public IP and at least port 443 open TCP/UDP. Nodes should have access to DNS, so they can autodiscover its own public IP. If your nodes can't have a public IP and/or the browsers in your infrastructure can't access to nodes, you will need to configure an external TURN server. You can configure a TURN server in the configuration file of the node at /opt/openvidu/.env. You can find more information about how to configure an external TURN server in the Allow users behind firewalls section.

Check cluster and nodes health 🔗

You can check the health of the cluster and the nodes by using the REST API healthcheck endpoint. If your cluster is working correctly it should return a response like this:

{
  "clusterStatus": "UP",
  "nodes": {
    "master_10.5.0.6": {
      "status": "UP",
      "id": "master_10.5.0.6",
      "uri": "http://10.5.0.6:5443",
      "disconnectedMediaNodes": []
    },
    "master_10.5.0.7": {
      "status": "UP",
      "id": "master_10.5.0.7",
      "uri": "http://10.5.0.7:5443",
      "disconnectedMediaNodes": []
    }
  }
}

If your status is UNSTABLE or DOWN it means that your cluster is not working correctly.

On the other hand there is an individual HTTP endpoint per node that you can use to check the health of a specific node and make your Load Balancer or Cloud Provider notice which nodes are working correctly and which ones are not.

For example, if you execute a GET request to http://10.5.0.6:4443/openvidu/health, if this node is working correctly it should return a 200 OK response. You can use this endpoint to autoconfigure your Load Balancer or remove unhealthy nodes from your Cloud Provider.