DIBS: Just-in-time Congestion Mitigation for Data Centers

Data centers must support a range of workloads with dif-fering demands. Although existing approaches handle rou-tine traffic smoothly, intense hotspots–even if ephemeral– cause excessive packet loss and severely degrade perfor-mance. This loss occurs even though congestion is typi-cally highly localized, with spare buffer capacity at nearby switches. In this paper, we argue that switches should share buffer capacity to effectively handle this spot congestion without the monetary hit of deploying large buffers at in-dividual switches. Specifically, we present detour-induced buffer sharing (DIBS), a mechanism that achieves a near lossless network without requiring additional buffers at in-dividual switches. Using DIBS, a congested switch detours packets randomly to neighboring switches to avoid dropping the packets. We implement DIBS in hardware, on software routers in a testbed, and in simulation, and we demonstrate that it reduces the 99th percentile of delay-sensitive query completion time by up to 85%, with very little impact on other traffic. 1