Virtual Cluster with Kubernetes

You will be using cloudmesh and ansible to build a kubernetes swarm cluster on cloud resources.

This may include one or more clouds while using cloudmesh as a basis for deployment. You will be developing a benchmark suite (you can resue benchmarks developed by others) and conduct a sophisticated set of benchmarks on such a cluster.

You will contrast these benchmarks with other virtual clusters set up for example on openstack. Integrating heterogeneous resources from multiple clouds would also be useful and identifying the impact on the benchmarks.

Develop carefully a plan for the benchmarks.

Instead of just running the becnhmark, on an existing infrastructure, we expect that the benchmarks can be replicated with a script or series of script to automatically gather the results in a table that than gets automatically plotted.