Default workflows, problem with CPU/memory usage

[dragoslav]

"... it is running ok for like 2-3 hours then gradually getting a lot of cpu time, then they (the machine) starts swapping. ... If I restart both dockers in marathon it looks ok for a couple of hours before the issue begins happening again."

[dragoslav]

This issue should be resolved by #813, but still requires further testing/confirmation.

[olafmol]

first we need to add metric collection to workflows, then we should create a stress-test environment and scripts @iamtew

[iamtew]

metricbeat is available in the workflow agent, see: magneticio/vamp-workflow-agent#10

gateway agent is almost there: https://github.com/magneticio/vamp-gateway-agent/tree/metricbeat

vamp container itself is planned, but not yet implemented: magneticio/vamp-docker-images#21