hesela.devProtocol-aware recovery
Protocol-aware recovery uses a distributed system's replication rules to recover damaged storage without violating its guarantees.
CTRL separates local fault detection from distributed recovery. It distinguishes missing, damaged and intact log entries; node-specific metadata instead uses local redundancy. Recovery assumes detectable corruption and may sacrifice availability to preserve committed data. [1]
This is a research approach, not a claim that replication alone repairs corruption. The companion analysis compares it with historical CORDS fault-injection evidence. [2]