When we started discussing moving DPDK to the Linux Foundation we talked about the possibility of establishing a DPDK lab. We agreed at the time that CI should be done in a distributed way, and Thomas has done great work to put this in place. We also said that it would be good to have a public lab for performance testing, so that we have an open and repeatable way of identifying any performance regressions caused by new contributions to DPDK.

Attached is a first draft of a proposal to define the scope and purpose of a DPDK lab. Once we've agreed on those, we can progress to the next level of detail and determine specific hardware requirements, get a more accurate estimate of hosting costs etc.

Please review and comment on this. For now I think moving@dpdk.org is the most appropriate mailing list for discussion on this because the proposal has been created in response to discussion there and because it's closely linked to discussions on budget etc., but I've copied ci@dpdk.org as well. We can discuss this during next Tuesday's meeting as well and agree on the next steps for progressing this.


Tim