grav
A collection of performance visualization tools for analyzing process execution, scheduler behavior, CPU tenancy, and JVM heap allocations.
What is grav?
Grav is a collection of performance visualization tools for Linux systems that helps analyze process execution, CPU scheduling, and JVM memory allocation. It generates visual reports like scheduler profiles, CPU tenancy charts, and flame graphs to identify performance bottlenecks and optimization opportunities. The tools integrate with existing Linux performance monitoring infrastructure like perf and BCC.
Performance engineers, DevOps professionals, and Java developers working on Linux systems who need to diagnose low-level performance issues, scheduler behavior, or JVM memory allocation patterns.
Grav provides ready-to-use scripts that simplify the generation of advanced performance visualizations without requiring deep expertise in Linux profiling tools. Its focus on visualization makes complex performance data more accessible and actionable compared to raw metric collection.
Overview
Performance visualisation tools
Use Cases
Best For
- Identifying kernel scheduler pre-emption issues in multi-threaded applications
- Analyzing CPU affinity and tenancy for performance optimization
- Visualizing JVM heap allocation patterns and hotspots
- Creating annotated flame graphs with thread names for Java applications
- Generating animated flame graphs to show performance changes over time
- Setting up a reproducible performance analysis environment with Vagrant
Not Ideal For
- Windows or macOS developers needing cross-platform performance analysis tools
- Projects using non-JVM languages without Java integration requirements
- Environments where installing kernel-level tools like BCC is restricted due to security policies
Pros & Cons
Pros
Generates SVG charts like flame graphs and scheduler profiles that transform raw performance data into easily interpretable visuals, helping identify bottlenecks quickly.
Includes specialized features such as heap allocation flamegraphs and thread-annotated flamegraphs, leveraging DTrace probes for deep Java application analysis.
Offers a Vagrant box that pre-installs all dependencies like perf and BCC, ensuring a consistent environment for performance testing on non-Linux machines.
Builds on established Linux tools like perf and iovisor BCC, avoiding redundancy and ensuring compatibility with existing performance monitoring ecosystems.
Cons
Requires manual cloning and configuration of multiple external repositories (e.g., perf-map-agent, flamegraph), which can be tedious and error-prone for new users.
Tied to Linux systems and primarily focused on Java, with no built-in support for other operating systems or programming languages, reducing versatility.
Produces static SVG files without interactive features or real-time monitoring, limiting dynamic performance investigation compared to live dashboards.
Frequently Asked Questions
Related Projects
leakcanaryA memory leak detection library for Android.
ZipkinZipkin is a distributed tracing system
async-profilerSampling CPU and HEAP profiler for Java featuring AsyncGetCallTrace + perf_events
BTraceProduction-safe dynamic tracing and diagnostics for Java applications—attach to live JVMs with no restart or recompilation.
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.