On (04/22/16 20:23), marcelo.leit...@gmail.com wrote: > My tests results were very similar to what I had without it. Varying to > better or worse, tending worse. Thing is, SCTP always works on > linearized skbs as it can't crawl on fragments, so those clone/trim
sorry to hear that. For RDS-TCP, the rx side does show a noticeable benefit with rds-stress. To an extent, this is also impacted by the packet size, and the type of test (for our DB workloads, we use request-response tests, and the packet size tends to typically be 8K req, 256 byt responses), so I guess ymmv. --Sowmini
