enable performance tests analisys in the pipeline
A certain level of performances through sdapi should be guaranteed from all CSP implementations. There are already simple program examples (pwrite pread) and a more complete one here (evaluate read-write perf. on different combinations of thread numbers and blocks sizes) that can be used in the pipeline to guarantee library performances. This last can be built using the provided build script (passing the --build-ptest options…. see README.md). both AWS/IBM should start looking into this example to understand if these works for them and how we can generally enable them in the pipeline (define acceptance level and execution context). I’ve created a bug here