Investigate the efficiency of autoscaling strategies when simulation software goes to the cloud
Wind, waves and currents shape our delta. Climate change and rising sea levels add further to this pressure. Simulation software exists that can simulate storm surges, hurricanes, tsunamis, detailed flows and water levels, waves, sediment transport and morphology, water quality and ecology, and is capable of handling the interactions between these processes.
The D-HYDRO Suite 1D2D is an example of such simulation software and is capable of simulating the processes described above and in recent years the engines of this suite can run large computations in parallel on distributed memory.
The era of cloud computing has just begun and we are not yet able to fully exploit its potential. Cloud computing offers horizontal scalability and elasticity and problems could be solved more in parallel. Many applications are not yet able to fully utilize the possibilities offered by the cloud, and even if they do, cloud platforms offer many configuration options, which leaves new users with choices that are not always easy to make. How to move to the cloud including reaping the full benefits of the cloud?
What will change if an application, or part of it, is migrating from the premises to the cloud? Should a software architect rethink design? Which viewpoints and quality attributes are touched upon?