Pinpoint APM Platform

This open-source Application Performance Management (APM) tool provides large-scale distributed system monitoring with a focus on helping developers analyze the overall structure of distributed systems and identify performance bottlenecks across microservice interactions. Originally developed by Jongho Kim and the Naver engineering team in 2012 as an internal tool for monitoring Naver's massive infrastructure, Pinpoint was donated to the Apache Software Foundation in 2021 and graduated as a top-level project. Key features include: distributed transaction tracing (inspired by Google Dapper, automatically tracing RPC calls, HTTP requests, database queries, and cache operations across service boundaries with minimal overhead), code-level visibility (showing exact method calls, arguments, and return values within a single transaction for debugging), service topology (real-time visualization of inter-service dependencies and call volumes without code changes), bytecode instrumentation (Java agent automatically injecting trace code into supported libraries including Spring, MyBatis, gRPC, Thrift, HTTP, Redis, MongoDB, and Kafka without source code modification), interactive call tree (detailed tree visualization of the full execution path with timing breakdown per method call), inspector (per-server monitoring of CPU, memory, GC, TPS, response time, and error rate with historical trend charts), agent statistics (per-agent runtime metrics including JVM, connection pool, and active thread counts), alarm (configurable threshold-based alerts on response time, error rate, and resource usage with email, SMS, webhook, and Slack notifications), modular agent (pluggable instrumentation modules for adding new framework support without code changes), and HBase storage (optimized for large-scale time-series trace and metric data with HBase and optional Elasticsearch backend).