SCALER is a collaborative project funded by the French National Research Agency (ANR).

The goal of the SCALER project is to deliver a software framework to enable smart microservice scaling capabilities going beyond the simple model of managing one microservice at a time, but rather considering cohorts of microservices based on their interaction patterns. Its research objectives are the following:

  1. Automatic Microservice Characterization. The first challenge of the SCALER project is to iden- tify and to propose mechanisms for harvesting relevant metrics at different (system, infrastruc- ture, application) levels. Metrics are to characterize single but also groups of microservices. Used for pre-deployment profiling or monitoring at runtime, metrics are to allow for a timely mapping between business-level Key Performance Indicators (KPI) and lower-level metrics.
  2. Microservice Interaction Patterns. If microservices are independent deployment entities, their runtime interactions strongly impact applications’ KPI. The second challenge of the SCALER project is to provide solutions for identifying microservice interactions patterns and for defining groups of microservices to be managed consistently vis-à-vis performance constraints. Typically, if an application service is composed of multiple microservices, the latter are to respond together to the service latency (global response time) or locality (multi-cluster) constraints.
  3. Microservice Smart Scaling. The third SCALER challenge is to design smart management strategies which take into account the previously defined microservice interaction patterns. SCALER is to define solutions for automated, timely and consistent coordinated scaling of groups of microservices.