#distributed-systems
Wiki 14
- Byzantine Fault Failure mode where a node produces wrong output while appearing functional
- Distributed Consensus Getting multiple processes to agree on a value: Paxos, Raft, PBFT, and the tradeoff space
- FLP Impossibility Proven 1985 result: no deterministic consensus in async systems with even one crash
- Linearizability The consistency model CAP calls C: an operation sees state at least as new as any operation that completed before it started
- LLMs are breaking 20-year-old system design Zknill on how stateful agents break the load-balancer-plus-database cloud-native assumption — the missing primitive is a routable transport name, and pub/sub channels fit
- Meerkat, Cloudflare's QuePaxa consensus service Cloudflare swaps Raft for QuePaxa: no required leader, no timeouts, global control-plane state
- Mesh LLM: Distributed Inference over iroh Pools GPUs across machines as one OpenAI-compatible API; splits big models across nodes over iroh's p2p QUIC transport
- Message Passing Is Shared Mutable State Tu et al's 2019 Go bug study confirms Edward Lee's 2006 prediction — channels are concurrent queues with all the bugs of shared state
- Message Passing vs Shared Memory The two camps of concurrency coordination, the case that they fail for the same reason, and what each actually buys you
- Multi-agentic Software Development is a Distributed Systems Problem Multi-agent LLM coding is a distributed consensus problem subject to FLP and Byzantine impossibility
- Please stop calling databases CP or AP Kleppmann (2015): CAP's definitions are too narrow to classify real databases, most of which are neither CP nor AP, so drop the labels
- Stateful agent routing primitive The missing piece in cloud-native architecture for talking to a specific durable process — a routable transport name that isn't a server, well-fit to pub/sub channels
- What Even Are Microservices? var0.xyz: microservices solve an organizational problem, and you pay distributed-system costs
- zknill.io Zach Knill's blog — distributed systems, durable execution, agent infrastructure, the routing primitive missing from cloud-native
Books 5
- Concurrency: The Works of Leslie Lamport Annotated reader of Lamport's papers — clocks, Byzantine Generals, Paxos, TLA+ — with commentary
- Database Internals: A Deep Dive into How Distributed Data Systems Work Petrov's two-part survey of storage engines and the distributed side of database systems
- Designing Data-Intensive Applications (2nd ed.) Kleppmann on the tradeoffs behind storage engines, replication, and stream processing
- Replication: Theory and Practice Survey volume on primary-backup, state-machine replication and atomic broadcast
- The Art of Multiprocessor Programming Herlihy and Shavit on linearizability, lock-free structures, and the consensus hierarchy