Network Reliability & Fault Tolerance Lab
Explore how network topology affects reliability under failure. Choose from star, ring, mesh, and tree topologies, inject node and link faults, and observe how packet delivery rates change. Discover why redundant paths are critical for fault-tolerant networks.
Guided Experiment: Redundancy vs Reliability
How does the number of redundant paths in a network affect its ability to deliver packets when nodes or links fail?
Write your hypothesis in the Lab Report panel, then click Next.
Network Topology
Controls
Highly interconnected graph. Many redundant paths.
Reliability Metrics
Data Table
(0 rows)| # | Trial | Topology | Nodes Failed | Links Failed | Packets Sent | Delivered | Lost | Delivery Rate % | Avg Latency ms |
|---|
Reference Guide
Fault Tolerance
A fault-tolerant network continues to operate when components fail. The key measure is how many simultaneous failures the network can absorb before losing connectivity between two endpoints.
Single points of failure, such as the central hub in a star topology, make the entire network vulnerable to one component's failure.
Redundancy & Independent Paths
Two paths between nodes are independent (edge-disjoint) if they share no links. The number of independent paths equals the minimum edge cut by the max-flow min-cut theorem.
A mesh network with k independent paths between two nodes can tolerate k-1 link failures and still maintain connectivity.
Packet Loss & Retransmission
Each link in a network has a probability p of dropping a packet. For a path with n hops, the probability of successful delivery is (1-p)^n, assuming independent link failures.
Shorter paths are more reliable because the packet crosses fewer lossy links. Rerouting through an alternate path can recover from loss on the primary path.
Network Availability
Network availability measures the fraction of node pairs that can communicate. A fully connected network has 100% availability. Faults reduce this metric.
High availability networks use redundant paths, backup nodes, and automatic failover to keep availability above 99.9% even during component failures.