Active-Passive Replication

A high-availability and load-balancing architecture where multiple nodes simultaneously process requests. Unlike the primary-standby model of active-passive, all replicas are active, distributing the workload and providing inherent redundancy. This pattern improves resource utilization and horizontal scalability but introduces complexity for state synchronization and conflict resolution across nodes handling concurrent writes.

• Key Benefit: Maximizes infrastructure investment by using all nodes for throughput.
• Key Challenge: Requires sophisticated consensus or eventual consistency mechanisms for data integrity.

A fundamental fault tolerance technique where a deterministic service is replicated across multiple machines. Each replica processes the same sequence of requests in the same order, ensuring they all undergo identical state transitions and produce the same outputs. Active-Passive Replication is one implementation strategy for SMR, where the primary sequences the requests for the passive replicas.

• Core Principle: Determinism is required; same input + same state = same output.
• Implementation: Often underpinned by consensus protocols like Raft or Paxos to agree on the request order.

The automatic process of switching to a redundant or standby system when the currently active component fails. This is the critical event that Active-Passive Replication architectures are designed to execute. A successful failover requires:

• Failure Detection: Using health checks, heartbeats, or timeouts.
• Leader Election: Promoting a passive replica to become the new active primary.
• Traffic Redirection: Updating load balancers or service discovery to route requests to the new primary.
• State Reconciliation: Ensuring the new primary has a sufficiently recent state to resume service.

A design pattern that prevents cascading failures by detecting remote service failures and failing fast. When a downstream agent or service (like a passive replica attempting to sync) fails repeatedly, the circuit breaker "trips" and blocks further calls for a period. This allows the failing component time to recover and prevents resource exhaustion in the calling system.

• Three States: Closed (normal operation), Open (failing fast), Half-Open (probing for recovery).
• Use Case: Protects an active primary from being overwhelmed by retries to an unhealthy passive replica.

A catastrophic failure condition in high-availability clusters where a network partition causes independent sub-clusters to believe they are the sole active group. In an Active-Passive system, this could result in multiple nodes assuming the active primary role, leading to data corruption from conflicting writes. Prevention mechanisms include:

• Quorums: Requiring a majority of nodes to agree on leadership.
• Fencing: Using STONITH (Shoot The Other Node In The Head) to forcibly disable a disputed primary.
• Careful Timeout Configuration: To distinguish network delay from true node failure.

A periodic probe or request sent to a service or agent to verify its operational status and readiness. In Active-Passive Replication, health checks are the primary mechanism for the orchestrator or load balancer to detect a failing active primary. Effective health checks:

• Are Liveness Probes: Determine if the process is running and responsive.
• Are Readiness Probes: Determine if the agent can handle requests (e.g., has synced state).
• Must be Lightweight: To avoid creating their own failure load.
• Must have Configurable Timeouts & Thresholds: To avoid false positives from transient issues.