Installatron Server Clustering: High Availability and Scalability

(revised 2025-03-05)

1. Introduction
      1.1. High Availability
      1.2. Scalability

2. Implementing
      2.1. Advanced Load Balancing Options

1. Introduction

Installatron Server enables high availability and scalability by allowing multiple Installatron Server nodes to be clustered together. By leveraging a cluster of Installatron Server nodes, businesses can ensure that their Installatron Server deployment remains reliable, responsive, and ready to scale as required:

1.1. High Availability

High Availability (HA) refers to a Installatron Server's capability to remain operational and accessible even in the face of hardware failures, network issues, or other disruptions by distributing the API and UI requests across a cluster of nodes. In a high-availability configuration, redundancy and failover mechanisms help to ensure that services remain online with minimal downtime.

• Redundancy
  In the event of a node failure, other nodes seamlessly take over processing requests, maintaining uninterrupted service.
• Efficient Resource Utilization
  Workloads are balanced across available nodes, optimizing system performance.

1.2. Scalability

Scalability refers to Installatron Server's capability to handle an increase in API and UI requests by distributing the requests across a cluster of nodes. As demand increases, additional nodes can be added to the cluster to balance the load efficiently. This ensures that no single node is overwhelmed, maintaining optimal performance and responsiveness.

• Load Balancing
  Requests are distributed across multiple nodes, preventing bottlenecks and helping to ensure no single node is overloaded.
• Seamless Expansion
  Additional nodes can be integrated as needed, with the cluster automatically adjusting to balance new workloads efficiently.

2. Implementing

Start by configuring a single Installatron Server node following the Typical installation instructions. While not required, you may choose to offload the MySQL/MariaDB database to a dedicated database server by including the necessary installation parameters.

Next, add your first additional Installatron Server node following the Additional Node installation instructions which instruct Installatron Server on the additional node to share the same database as the first Installatron Server node.

The first and second Installatron Server nodes are now clustered. They now share the same database and workload, helping to ensure the Installatron Server cluster as a whole remains dependable, efficient, and scalable to accommodate increasing demand.

Additional nodes can be added as demand requires by repeating the Additional Node step.

To remove a node, simply power off the node and the cluster will adapt to the removed node by consolidating requests to the remaining node(s).

2.1. Advanced Load Balancing Options

By default, requests are assigned to available nodes in a Round Robin manner, distributing traffic sequentially to help maintain an even load. If a node becomes unavailable, the remaining nodes automatically handle the requests to ensure continuity.

A dedicated load balancer, though optional, serves as an intermediary that optimizes request distribution more efficiently. Load balancers can be configured in various ways to meet specific infrastructure needs and performance goals.

Common load balancing configurations include:

• Round Robin
  Requests are distributed sequentially across all available nodes, ensuring even distribution of traffic.
• Least Connections
  Directs traffic to the node with the fewest active connections, optimizing resource utilization during peak loads.
• Weighted Load Balancing
  Assigns different weights to nodes based on their processing power or capacity, ensuring that more powerful nodes handle a larger share of the requests.