# dgop ### Stateless, cursor-based system and process monitoring [![Documentation](https://img.shields.io/badge/docs-danklinux.com-9ccbfb?style=for-the-badge&labelColor=101418)](https://danklinux.com/docs/dgop) [![GitHub Release](https://img.shields.io/github/v/release/AvengeMedia/dgop?style=for-the-badge&labelColor=101418&color=a6da95)](https://github.com/AvengeMedia/dgop/releases) [![Arch Linux](https://img.shields.io/archlinux/v/extra/x86_64/dgop?style=for-the-badge&labelColor=101418&color=f5a97f)](https://archlinux.org/packages/extra/x86_64/dgop/) [![GitHub License](https://img.shields.io/github/license/AvengeMedia/dgop?style=for-the-badge&labelColor=101418&color=b9c8da)](https://github.com/AvengeMedia/dgop/blob/master/LICENSE)

---

System monitoring tool with CLI and REST API built in Go. Fast, single binary, JSON output, OpenAPI spec. Use standalone for system monitoring, or as a companion for [DankMaterialShell](https://github.com/AvengeMedia/DankMaterialShell) to unlock system information widgets. **[Full documentation →](https://danklinux.com/docs/dgop)** --- ## Installation ### Latest Release Download the latest binary from [GitHub Releases](https://github.com/AvengeMedia/dgop/releases/latest) ### Arch Linux ```bash # Using pacman sudo pacman -S dgop ``` ### Build from Source ```bash # Build it make # Install system-wide sudo make install # Or just run locally go run ./cmd/cli [command] ``` ## Basic Commands ```bash # See all at once dgop all # Just CPU info dgop cpu # Memory usage dgop memory # Network interfaces dgop network # Disk usage and mounts dgop disk # Running processes (sorted by CPU usage) dgop processes # System load and uptime dgop system # Hardware info (BIOS, motherboard, etc) dgop hardware # GPU information dgop gpu # Get temperature for specific GPU dgop gpu-temp --pci-id 10de:2684 # List available modules dgop modules ``` ## Meta Command Mix and match any modules you want: ```bash # Just CPU and memory dgop meta --modules cpu,memory # Everything except processes dgop meta --modules cpu,memory,network,disk,system,hardware,gpu # GPU with temperatures dgop meta --modules gpu --gpu-pci-ids 10de:2684 # Multiple GPU temperatures dgop meta --modules gpu --gpu-pci-ids 10de:2684,1002:164e # Everything (same as 'dgop all') dgop meta --modules all ``` ## JSON Output Add `--json` to any command: ```bash dgop cpu --json dgop meta --modules gpu,memory --json ``` ## Process Options ```bash # Sort by memory instead of CPU dgop processes --sort memory # Limit to top 10 dgop processes --limit 10 # Skip CPU calculation for faster results dgop processes --no-cpu # Combine options dgop meta --modules processes --sort memory --limit 20 --no-cpu ``` ## API Server Start the REST API: ```bash dgop server ``` Then hit these endpoints: - **GET** `/gops/cpu` - CPU info - **GET** `/gops/memory` - Memory usage - **GET** `/gops/network` - Network interfaces - **GET** `/gops/disk` - Disk usage - **GET** `/gops/processes?sort_by=memory&limit=10` - Top 10 processes by memory - **GET** `/gops/system` - System load and uptime - **GET** `/gops/hardware` - Hardware info - **GET** `/gops/gpu` - GPU information - **GET** `/gops/gpu/temp?pciId=10de:2684` - GPU temperature - **GET** `/gops/modules` - List available modules - **GET** `/gops/meta?modules=cpu,memory&gpu_pci_ids=10de:2684` - Dynamic modules API docs: http://localhost:63484/docs ## Examples ### Get GPU temps for both your cards ```bash dgop meta --modules gpu --gpu-pci-ids 10de:2684,1002:164e ``` ### Monitor system without slow CPU calculations ```bash dgop meta --modules cpu,memory,network --no-cpu ``` ### API: Get CPU and memory as JSON ```bash curl http://localhost:63484/gops/meta?modules=cpu,memory ``` ### API: Get GPU with temperature ```bash curl "http://localhost:63484/gops/meta?modules=gpu&gpu_pci_ids=10de:2684" ``` ## Real-time Monitoring with Cursors dgop supports cursor-based sampling for building real-time monitoring tools like htop. Instead of relying on instantaneous snapshots, you can track system state changes over time for more accurate CPU usage calculations and network/disk rates. The cursor system works by: - Taking an initial measurement that establishes baseline metrics and timestamps - Returning a base64-encoded cursor containing the current state data - Using that cursor in subsequent calls to calculate precise percentages and rates over the sampling interval This approach accounts for the actual time elapsed between measurements, making it ideal for monitoring tools that poll every few seconds. ### CPU Usage with Cursors ```bash # First call - establishes baseline and returns cursor dgop cpu --json # Returns: {"usage":1.68, ..., "cursor":"eyJ0b3RhbCI6WzE2MjMwLjAzLDUuOTUsNTEyMy4yNV0..."} # Wait a few seconds, then use cursor for accurate CPU calculations sleep 3 dgop cpu --json --cursor "eyJ0b3RhbCI6WzE2MjMwLjAzLDUuOTUsNTEyMy4yNV0..." # Returns more accurate usage percentages based on time delta ``` ### Process Monitoring with Cursors ```bash # First call - establishes process baseline dgop processes --json --limit 5 # Returns: {"processes":[...], "cursor":"W3sicGlkIjoyODE2NTYsInRpY2tzIjozOS43Mix9XQ..."} # Use cursor for accurate per-process CPU calculations sleep 2 dgop processes --json --limit 5 --cursor "W3sicGlkIjoyODE2NTYsInRpY2tzIjozOS43Mix9XQ..." ``` ### Network Rate Monitoring ```bash # First call - establishes network baseline dgop net-rate --json # Returns: {"interfaces":[...], "cursor":"eyJ0aW1lc3RhbXAiOiIyMDI1LTA4LTExVDE2OjE1OjM1..."} # Get real-time transfer rates sleep 3 dgop net-rate --json --cursor "eyJ0aW1lc3RhbXAiOiIyMDI1LTA4LTExVDE2OjE1OjM1..." # Returns: {"interfaces":[{"interface":"wlp99s0","rxrate":67771,"txrate":16994}]} ``` ### Disk I/O Rate Monitoring ```bash # Establish disk I/O baseline dgop disk-rate --json # Returns cursor for disk rate calculations # Get real-time disk I/O rates sleep 2 dgop disk-rate --json --cursor "eyJ0aW1lc3RhbXAiOiIyMDI1LTA4LTExVDE2OjE2..." ``` ### Combined Monitoring with Meta Command ```bash # Monitor CPU, processes, and network rates together dgop meta --modules cpu,processes,net-rate --json --limit 10 # Use multiple cursors for comprehensive monitoring dgop meta --modules cpu,processes,net-rate --json --limit 10 \ --cpu-cursor "eyJ0b3RhbCI6WzE2MjMwLjAz..." \ --proc-cursor "W3sicGlkIjoyODE2NTYsInRpY2tzIjo..." \ --net-rate-cursor "eyJ0aW1lc3RhbXAiOiIyMDI1LTA4LTEx..." ``` ## Development ```bash # Build make # Run tests make test # Format code make fmt # Build and install make && sudo make install # Clean build artifacts make clean ``` ## Requirements - Go 1.22+ - Linux (uses `/proc`, `/sys`, and system commands) - Optional: `nvidia-smi` for NVIDIA GPU temperatures ## Why Another Monitoring Tool? Because nothing did what i wanted, i didnt want to run a metrics server, I wanted GO because its fast and compiles to a single binary, bash scripts got too messy. TL;DR single binary cli and server with json output, openapi spec, and a bunch of data.