uber-common/jvm-profiler

Health v2 · maintenance only

Flags: no_releases no_license

Adoption not part of the score

1804 stars · 341 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

A Java Agent library from Uber that collects CPU, memory, IO metrics and stacktraces from JVM processes in a distributed way, with reporters for Kafka, console, file, Redis, and InfluxDB. It also traces arbitrary Java methods and arguments without code changes, originally built for profiling Spark applications across many machines.

Use cases

• profile CPU and memory metrics of Spark executors across a cluster
• trace HDFS name node call latency for Spark applications
• identify hot files read or written by Spark jobs
• attach a javaagent to any JVM process to collect metrics
• ship JVM metrics to Kafka for downstream analysis
• trace method durations and arguments without modifying user code

When to choose

• you need distributed metrics collection for dozens or hundreds of JVM processes in one application
• you want method-level tracing of JVM code with zero code changes
• you run Spark or Hadoop workloads and need to correlate metrics across executors

When to avoid

• you need a full interactive profiler UI with flame graphs
• you profile non-JVM languages or native code
• you need continuous sampling-style CPU profiling rather than interval metrics

Facets

library · maturity maintenance

monitoring tracing benchmarking performance developer-tools big-data jvm cross-platform java-agent spark hadoop kafka-reporter jvm-metrics method-tracing

Member repositories

For agents

markdown · JSON · MCP: product_card(name="uber-common/jvm-profiler")