I implemented Raft, a KV store, and a sharded system in Go…
0 reactions
I implemented Raft, a KV store, and a sharded system in Go
I recently completed the labs from MIT 6.5840 Distributed Systems and implemented everything in Go, including:
* Raft consensus algorithm
* A replicated Key/Value store
* A sharded KV system with dynamic reconfiguration
The implementation focuses a lot on concurrency and failure handling:
* goroutines for RPC handling and background tasks
* channels for coordination between Raft and the state machine
* dealing with unreliable networks (dropped / delayed / out-of-order RPCs)
Some interesting challenges:
* ensuring commitIndex never goes backward under out-of-order RPC responses
* handling retries safely with client/request IDs (idempotency)
* keeping deduplication state consistent across snapshots and shard transfers
I wrote a detailed README explaining both the design and the tricky edge cases I encountered.
#programming #technology #dev
I recently completed the labs from MIT 6.5840 Distributed Systems and implemented everything in Go, including:
* Raft consensus algorithm
* A replicated Key/Value store
* A sharded KV system with dynamic reconfiguration
The implementation focuses a lot on concurrency and failure handling:
* goroutines for RPC handling and background tasks
* channels for coordination between Raft and the state machine
* dealing with unreliable networks (dropped / delayed / out-of-order RPCs)
Some interesting challenges:
* ensuring commitIndex never goes backward under out-of-order RPC responses
* handling retries safely with client/request IDs (idempotency)
* keeping deduplication state consistent across snapshots and shard transfers
I wrote a detailed README explaining both the design and the tricky edge cases I encountered.
#programming #technology #dev
| 2,000 mlx | |
| 2 | |
| 0 |