partition-failures-compound-across-membership-and-consensus

Status: IN

Network partitions create cascading failures across the membership and consensus layers: gossip-based failure detection is the single correctness gate for replication, and Raft's partition hazard (stale leader silently accepting doomed writes, forced re-election on rejoin) depends on membership accuracy that gossip's timeout-based suspicion may misjudge.

Justifications

inaccurate membership detection feeds incorrect state to consensus, compounding both layers' partition vulnerabilities

Depends on (SL): gossip-failure-detection-governs-cluster-correctness, raft-partition-creates-dual-hazard

Depended on by

JSON