AI Briefing#distributed-systems
The latest AI and developer news about #distributed-systems, with the original source and a short summary.
The latest AI and developer news about #distributed-systems, with the original source and a short summary.
A collection of 56 software principles spanning teams, architecture, quality, and design.
This covers how Airbnb built a metrics storage system that is resilient to failures.
Cloudflare redesigned Workflows to achieve 50,000 concurrency and 3x creation speed.
A Datomic-compatible Clojure Datalog DB that lets you freely query past states.
The more complex a multi-agent system gets, the more it should start from the simplest coordination pattern.
Using the Mean Girls analogy, it breaks down Raft's quorum and leader election in an easy-to-understand way.
Spotify pre-generated 1.4 billion reports based on listening data from 350 million people.
500ms Pod latency effectively paralyzed the service, while 250ms only increased latency but held up.
Workflow 4.1 Beta introduces an event sourcing architecture to strengthen workflow reliability and debugging performance.
We internalized an external Batch Scheduler using Quartz and a common library, calling it OnQuartz.
This covers the technical challenges and solutions Olive Young faced while building its own Kafka-based global logistics system (GMS) without relying on external solutions.
This presents a strategy for efficiently using local cache and Redis while considering data consistency in distributed systems.
Kakao Pay redesigned its RabbitMQ-based delayed transfer service using Kafka, improving processing speed by 8x.
KakaoPay shares how it resolved a concurrency issue encountered while operating a Slackbot in a distributed system environment, using transactions and Lock.
To resolve delays in session_id ingestion caused by traffic spikes, the team separated it out using Redis Stream.
ContributorService and NetworkQueryEngine connect multiple RAGs together like a network.
They switched from manual monthly settlement to daily automation, achieving 4x faster processing through idempotency and parallel processing.
Kakao applied a distributed Radix Tree to MetaKage to overcome the Ordered Listing limitations of LSMT and optimized performance using Bloom Filters.
70% of 2021 participants had no prior cloud experience, and the program supports career transitions through direct training by current colleagues.
Kakao DKOS detailed the etcd Raft algorithm, Learner state, compaction, and backup strategies to enable zero-downtime upgrades.
This explains how to leverage Redis and a key-renewal mechanism to limit the number of concurrent Websocket connections in a distributed environment.