Aemon: Information-agnostic Mix-flow Scheduling in Data Center Networks

Data center networks carry a mix of flows, some with deadlines and some without. Existing mix-flow transport designs assume prior knowledge of flow sizes, which may not hold in practice. Without such information, mix-flow scheduling becomes particularly challenging due to (1) the lack of precise rate control of deadline flows with minimal impact on non-deadline flows; (2) difficulty in assigning priority to both two types of flows. We present Aemon, an information-agnostic mix-flow transport. Aemon relies on a new congestion control mechanism based on urgency---the ratio between the elapsed transmission time and the remaining time to deadline---to strategically modulate the sending rate of deadline flows. In addition, Aemon uses a novel two-level priority scheduling policy to differentiate mix flows. As time goes, a deadline flow's priority level increases as its urgency rises, and a non-deadline flow's priority level decreases as it sends more bytes similar to PIAS to approximate Shortest-Job-First policy. While in the same priority level, non-deadline flows take precedence to avoid possible starvation caused by aggressive deadline flows. Extensive simulations on ns-2 show that Aemon outperforms existing information-agnostic schemes and is only slightly worse than state-of-the-art Karuna.