Distributed load generation¶
Some OpenStack projects (Marconi, MagnetoDB) require a real huge load, like 10-100k request per second for benchmarking.
To generate such huge load Rally have to create load from different servers.
- Rally can't generate load from different servers
- Result processing can't handle big amount of data
- There is no support for chunking results