Below

Detailed Introduction

Below is a time-traveling resource monitor for modern Linux systems designed to view and record historical system data. It supports live observation (live), continuous recording (record), and replaying historical snapshots (replay) for post-mortem analysis of performance events, resource usage trends, and cgroup hierarchies. Below exports scriptable outputs suitable for integration with monitoring stacks such as Prometheus and Grafana.

Main Features

• Multiple operation modes: live (real-time viewing), record (persistent collection), replay (playback of historical data), and dump (script-friendly exports like JSON/CSV/OpenMetrics).
• Records process and cgroup-level metrics, including Pressure Stall Information (PSI).
• Snapshot and replay functionality enables reproducible offline analysis and debugging.
• Implemented in Rust for performance and reliability; available as distribution packages and Docker images.

Use Cases

• Incident investigation: replay recorded data to locate transient or intermittent issues.
• Performance regression: compare replayed runs across time to identify regressions.
• Cluster and container monitoring: complement Prometheus/Grafana with event replay capabilities.
• Automated testing: capture snapshots for CI-based replay and validation.

Technical Features

• Written primarily in Rust with a focus on low overhead and robustness.
• Scriptable export formats include JSON, CSV, and OpenMetrics for downstream processing.
• Targets modern cgroup interfaces and kernel metrics (note: no support for cgroup v1).
• Distributed via Docker images and packaged for common Linux distributions for easy deployment and integration.