Configure API Portal for high availability

You can configure API Portal for active-active high availability (HA) deployment to ensure that there is no single of point of failure in the system. This helps to eliminate any potential system downtime in production environments.

HA in production environment

To achieve a API Portal HA deployment, you can deploy two API Portal instances (or nodes) behind a load balancer with a shared Network File System (NFS) storage configuration.

Note   API Portal only supports two-node HA deployment.

The HA deployment provides both high availability and horizontal scalability. The load balancer performs Transmission Control Protocol (TCP) checks at the network level on both active API Portal instances. If either of the crucial services (Apache or MySQL) becomes unavailable in one instance, the load balancer redirects traffic to the other, available node. When the first instance returns to normal operation, the traffic is again evenly distributed between the two instances.

The following diagram illustrates the API Portal HA setup:

The active-active high availability configuration

  • Users access the public IP of the load-balancer using a HTTP or HTTPS connection.
  • The load balancer spreads the connections between the two active API Portal instances on the Apache web servers.
  • Joomla! sends SQL queries to the database within its own API Portal instance.
  • Master-master replication is in place between the MySQL servers in the two instances, guaranteeing database consistency. This is direct communication between the two instances and does not involving the load balancer.
  • All static content (CSS, images, HTML files), and Joomla! PHP files are located on a shared NFS storage.

Related Links