Darwin and MigCast

Introduction to Darwin: NoSQL Schema Management Middleware

Introduction to MigCast: Tool-based Advisor Exploring Data Migration Strategies

We can now provide a Docker image of the Darwin prototype (including MigCast) together with a how-to:

A cost model behind the tool-based advisor MigCast takes several characteristics into account:​​ Data Set, Workload, Schema Evolution, DBMS and Cloud Pricing.​​

29 invariants automatically check the logs of all releases of schema evolutions for all runs of the Monte Carlo simulation. A traffic light system makes a quick assessment of a simulation run possible.

MigCastDB persists all simulation results to ensure reproducibility. The data model of MigCastDB includes all simulation parameters of a Monte Carlo-based probabilistic algorithm, its results and calculated statistical measures.

We recorded application data metrics of the simulation runs through the Grafana tool. View exemplary charts of CPU load, processes, and used memory in detail.

Contact: Uta Störl, University of Hagen. ✉ uta.stoerl@fernuni-hagen.de