A metrology agent for measuring electric power and energy consumption of tech services, enabling sustainability-focused decisions.
Scaphandre is a metrology agent designed to measure electric power and energy consumption of technology services and applications. It provides detailed metrics that help organizations and individuals understand the energy footprint of their systems, enabling data-driven decisions to improve sustainability. The agent supports various platforms and integrates with popular monitoring tools to streamline energy consumption analysis.
System administrators, DevOps engineers, and sustainability-focused tech teams who need to monitor and optimize the energy consumption of their infrastructure, including bare metal hosts, virtual machines, and Kubernetes clusters.
Developers choose Scaphandre for its lightweight design, multi-platform support, and seamless integration with existing monitoring stacks like Prometheus. Its unique ability to expose power metrics to virtual machines breaks the opacity of cloud and virtualized environments, providing transparency for sustainability initiatives.
⚡ Energy consumption metrology agent. Let "scaph" dive and bring back the metrics that will help you make your systems and applications more sustainable !
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Scaphandre aims for near-zero overhead in resource and power consumption, as stated in its philosophy, making it suitable for continuous monitoring without significant impact on system performance.
It works on GNU/Linux, Windows 10/11/Server, and Kubernetes, with packages for RHEL, Debian, and Windows, ensuring broad compatibility across diverse infrastructures.
Exposes metrics as a Prometheus HTTP exporter and supports push mode to Prometheus Push Gateway, seamlessly fitting into existing observability stacks for easy data collection.
Unique ability to expose power metrics from hypervisor to virtual machines, allowing VM-level energy tracking as if on bare metal, as detailed in the features for QEMU/KVM.
The project is in early development, as warned in the README, which means features may be incomplete, documentation sparse, and APIs subject to breaking changes.
Relies on specific power sensors (e.g., Intel RAPL) and hypervisor features; if unsupported, measurement is impossible, limiting its applicability to modern hardware only.
Requires configuration for different outputs and platforms, with separate installation guides for Linux and Windows, which can be cumbersome and time-consuming to implement.