Course contentsShow
← System Design
Lesson 1374 of 1,91934. Data Replication StrategiesFree lesson

Tree Replication Topology

Hierarchical replication where nodes replicate to child nodes; reduces load on leader but increases propagation delay.

Tree Replication Topology

You'll learn: How hierarchical tree-based replication distributes write load while introducing trade-offs in propagation delay.

In Tree Replication Topology, data replication follows a hierarchical structure similar to an organizational chart or family tree. The leader (root node) replicates changes to a few intermediate nodes, which in turn replicate to their own child nodes, and so on down the tree.

How It Works

Imagine a corporate memo distribution system. The CEO tells five VPs, each VP tells their five managers, and each manager tells their five team members. Instead of the CEO emailing 125 people directly, the message cascades down through levels.

In system design, the leader writes data once and sends it to a small number of first-tier replicas. These replicas become "mini-leaders" for their subtree, forwarding changes to second-tier replicas, which may forward to third-tier nodes, and so on.

The Core Trade-Off

Advantages:

  • Reduces leader load – The root node handles far fewer replication connections
  • Scales fan-out – Can reach thousands of nodes without overwhelming the leader
  • Network locality – Child nodes can be geographically grouped near their parent

Disadvantages:

  • Increased propagation delay – Changes must traverse multiple hops; a 3-level tree means 3 sequential replication steps
  • Intermediate node failure risk – If a middle node fails, its entire subtree loses updates until recovery
  • Complexity – Requires managing the tree structure and handling topology changes

Key Takeaway: Tree replication topology trades leader scalability for increased replication latency—ideal when you need to reach many replicas but can tolerate multi-hop propagation delays.