BABEL
Big dAta BEnchmarking pLatform, a generic, architecture-level and scalable benchmarking platform, destined to the real-time collection, analytics and reporting of data and events shared between the layers of complex Big Data applications.
Benchmark name is inspired from the myth of the Tower of BABEL, used to justify the origin of the presence of multiple spoken languages.
Features
The main features of BABEL:
- Genericity: technology-, application- and architecture-agnostic.
- Dynamicity: runs in parallel with the system, without interfering with its behavior, and with negligible overhead.
- Adaptability: adapts to any number of layers in the Big Data application, as well as to any type of data and workloads.
- Multi-Paradigm: able to benchmark streaming and batch processing systems, as well as operational and decisional platforms.
- End-to-End Solution: evaluates the behavior of the whole system, by measuring application- and system-level metrics.
- Real-time Reporting: offers a real-time reporting layer, that illustrates the system’s metrics on the fly.
- Distribution and Scalability: configurable, and offer a scalable collection, processing engines and storage, in order to enable the continuous archiving of metrics for a better analysis.
- Agility: flexible during the different integration phases and also at run-time, new layers, metrics, and processing jobs can be added on the fly.
Benchmark Architecture
Technical Architecture
