» Tag
distributed-systems
62 postsThere Is No "Now" in Distributed Systems
Inspired by Einstein's relativity: NTP clock sync is an illusion. Learn how logical clocks, CAP theorem, and TrueTime solve event ordering in distributed systems.
Prefill/Decode Disaggregation Can Worsen Tail Latency, Not Fix It
Splitting prefill and decode across GPU pools adds queues and KV transfer overhead that can worsen tail latency without careful control-loop design.
Why AT Protocol Traded Local-First for Federated Scale
How AT Protocol replaced local-first p2p design with a federated PDS network to meet Twitter-scale replication demands.
CommitBrief — AI code reviews, right in your terminal
A provider-agnostic, local-first CLI that reviews your staged changes, a historic range, or a whole GitHub pull request. Zero telemetry, no server. Free and open source.
commitbrief.comMeta Rebuilds BLOB Storage to Cut GPU Stalls in AI Training
Meta redesigned its BLOB-storage architecture with unified metadata and direct-to-client streaming to eliminate GPU stalls in AI training.
OpenRaft: An async Rust Raft crate with key protocol improvements
OpenRaft brings generalized membership changes, low-conflict elections, and high-throughput batching to async Rust Raft consensus.
LLM Agent Performance Is a Distributed Systems Problem
FixBugs's team shows how token math, parallel chunking, and queues cut LLM agent latency: file analysis dropped from 10 minutes to 40 seconds.
Why Distributed Systems Are Often Slower Than One Laptop
The 2015 COST paper showed distributed systems often lose to a single thread. A worked case study reveals the real cost of premature scaling.
Rate Limiting at Scale: Choosing the Right Throttling Algorithm
Why in-memory rate limiting silently breaks under horizontal scaling, and how token bucket, sliding window and fixed window algorithms trade off in production.
Using Kubernetes ConfigMaps as a Real-Time State Store
A Python demo shows how Kubernetes' watch API turns ConfigMaps into a real-time pub/sub change feed using asyncio, without polling or pod restarts.
How Postgres-Backed Queues Scale to 30K Workflows per Second
DBOS details how SKIP LOCKED, conditional isolation levels, and partial indexes let Postgres queues scale to 30K workflows per second.