This article outlines Pinterest's transition from a batch-oriented database ingestion system to a real-time, unified framework using Change Data Capture and modern data processing technologies. It addresses the challenges faced with legacy systems and details the architectural improvements that led to lower latency and better resource efficiency.

Arc is a high-performance time-series database capable of ingesting 2.4 million metrics per second, along with logs, traces, and events using a unified MessagePack columnar protocol. Currently in alpha release, it features a stable core with ongoing developments, including advanced SQL analytics via DuckDB, flexible storage options, and built-in token-based authentication, making it suitable for development and testing environments. The system is designed for high-throughput ingestion, low latency, and efficient data management, aiming to support observability across various telemetry types.