Posts tagged: TiKV
- The Multi-Raft Architecture
2026-03-25 12:00 MDT
The Raft consensus protocol is widely adopted for building fault-tolerant distributed systems. It ensures that even if a node crashes or becomes unreachable, the cluster agrees on a single consistent state. However, single-group Raft does not scale well.
Read more → - Runtime Choices in Distributed Systems
2026-03-25 15:30 MDT
The Raft algorithm is language-agnostic, but the runtime underneath it is not. GC pauses, CPU scheduling, and memory models shape which implementation strategies are even possible.
Read more →