we're looking for a strategy to more evenly distribute build requests among our build agents. Benchmark may or may not be a good indicator - what we really need is a way to change/experiment with different strategies. That is, we need a pluggable mechanism to provide these strategies, i.e. via plugins or scripts (provided the API gives us any information about grid / agents etc. to work with).
The benchmark we find only indicates how good the machine is at start up. one idea could be to update the benchmark periodically.
thanks in advance