RECAP Monitoring and Storage Infrastructure

The RECAP Monitoring and Storage Infrastructure integrates a set of established storage, data processing, and data distribution tools in a common toolset designed for geodistributed infrastructures and the applications running on top of them. It provides a backend-agnostic, decentralised integration point for monitoring data from different layers. It recognises the need to treat this data differently depending on the follow-up usage scenarios (offline analysis, online decision making, and visualisation) and accordingly provides different storage back-ends and connection points. In RECAP, this is used as an integration and normalisation point between the data generation components (data analytics, machine learning, infrastructure optimisation, application optimisation, and visualisation).

Benefits: The RECAP Monitoring and Storage Infrastructure is suitable for geo-distributed environments including edge/cloud computing and its distributed nature supports IoT environments, and its generic approach integrates seamlessly into existing environments. As an integration point, it enforces the normalisation of metric names and data formats so that independent from the underlying monitoring stack, consuming applications (such as machine learning) can always assume the same structure of data and the same semantics when interpreting it. This ensures re-usability of optimisation algorithms and machine learning tools. It covers the entire stack ranging from hardware components (servers, power consumption etc.) over virtual levels (OpenStack, Containers) to application metrics. It stores data for only for a limited, configurable amount of time resulting in low resource requirements. For scenarios with long-term storage requirements, it can be integrated with more durable back-ends.

Licence: Open Source.


Contact: Ulm University, Germany | Dr. Jörg Domaschka | [email protected]