RAP in Spotify: an index for point queries
RAP in Spotify demonstrates how an external index on top of Parquet accelerates point queries in the data lake without copying data to serving databases
Cloud-Native
Cloud-Native on ThecoreGrid explores how to design, run, and scale resilient systems built for dynamic cloud environments.
We cover practical architecture patterns around containers, Kubernetes, service discovery, configuration management, autoscaling, and immutable infrastructure. The focus is on production realities: multi-cluster operations, reliability under failure, cost control, observability, and secure workload isolation. You’ll find deep technical analysis of platform engineering, GitOps, Infrastructure as Code, traffic management, rollout strategies, and day-2 operations in highload systems. Instead of basic tutorials, we break down trade-offs between portability and provider-native services, speed and governance, flexibility and operational complexity. Content is curated from BigTech practices, real incident post-mortems, and hard lessons from cloud migrations at scale. The Cloud-Native tag is built for architects, platform and backend engineers, DevOps teams, and SREs who need robust, maintainable, and scalable cloud infrastructure for mission-critical products.
Service Topology: how Netflix keeps the dependency map live
Netflix Service Topology: how the real-time service map maintains integrity under load, using backpressure, SSE, and three processing stages
QUIC load balancing without state and rewriting
Stateless QUIC Load Balancing: How the Data Plane Ensures PCC, Reduces Latency, and Blocks 0-RTT Attacks Without Server Modifications
Scalability in Data-Intensive Systems Without Illusions
Scalability in data-intensive systems: how to choose between horizontal and vertical scaling and avoid rising latency and costs
KEDA autoscaling by backlog in SQS
KEDA autoscaling by backlog in SQS: how to scale Kubernetes by queue instead of CPU, reducing delays and costs. –>
Disaggregated databases under the pressure of the cloud economy
Disaggregated databases are transforming cloud architecture: how decoupling compute and storage affects system scalability, cost, and fault tolerance
Cloudflare Workers Scales cdnjs to 9 Billion
Cloudflare Workers and R2 have become the foundation for cdnjs. The architecture handles 9 billion requests per day and changes the approach to CDN pipelines. The cdnjs system has hit a wall not in delivery, but in evolution. With 108,000 requests per second and a 98.6% cache hit rate, the delivery layer operated stably. Degradation … Read more
ThecoreGrid Radar: This Week’s Trends in High-Performance Computing and Architecture
A curated selection of architecture insights and research releases we explored this week. Infrastructure 🔹 Spanergy: Energy-Aware Distributed Tracing for MicroservicesIntroduces an energy-efficient approach to microservice monitoring that reduces the power overhead of distributed tracing, making it particularly relevant for cloud-native architectures.Read the paper (EN) 🔹 ProFlow: RL-Based Proactive Flow PlacementApplies reinforcement learning to optimize … Read more
controller-runtime cache as the source of truth
controller-runtime cache: how reads, watches, and the reconcile loop work in Kubernetes, and why this impacts latency, memory, and consistency –>
Satellite inference on continents without degradation
Satellite inference on terabytes of data: how the OlmoEarth platform works and the architectural solutions that eliminate I/O and scaling bottlenecks