Daily Edition · Page 4

THE DAILY INDEX

Technology, AI, software, security, crypto and business

ARTICLES & TUTORIALS

TLDR Web Dev

HOW CLOUDFLARE MITIGATES THOUSANDS OF DDOS ATTACKS EVERY HOUR

Cloudflare mitigates an average of 2,200 DDoS attacks per hour. It measures attack size using metrics like bits per second (BPS), requests per second (RPS), and packets per second (PPS), analyzing attack vectors, targets, and payload signatures to identify and fingerprint attacks. Cloudflare has multiple protection strategies, like rate limiting, caching, CDNs, reducing attack surface, continuous monitoring, and machine learning.

AUTOMATED ACCESSIBILITY TESTING AT SLACK

Slack implemented automated accessibility testing using Axe, integrating it with their Playwright E2E framework to identify WCAG violations. They had initial challenges integrating Axe directly into their React Testing Library, but were able to successfully implement a system where accessibility checks run alongside existing tests, reporting critical violations. The full process includes a custom Playwright fixture to simplify test creation and uses an environment flag to control test execution on-demand, scheduled, and with CI.

MISTAKES ENGINEERS MAKE IN LARGE ESTABLISHED CODEBASES

Working in large, established codebases is challenging, requiring skills not learned in smaller projects or open source contributions. The most important skill is maintaining consistency by following existing code patterns and avoiding isolated, "clean" implementations. Understanding the codebase's production usage, particularly high-traffic areas, is also really important, as is careful testing and monitoring due to the impossibility of exhaustive testing.

HOW AIRBNB BUILT A KEY-VALUE STORE FOR PETABYTES OF DATA

Airbnb's derived data infrastructure evolved from HFileService, a read-only batch processing system, to Nebula, which combined batch and real-time data but suffered from inefficiencies and complexity. To overcome these challenges, Airbnb developed Mussel, a scalable, low-latency key-value store using technologies like Apache Helix, Kafka, and HRegion to handle real-time and batch data seamlessly. Mussel now supports over 130TB of data with high availability and sub-8ms read latency.

TLDR IndexPage 4