TY - GEN
T1 - Programmable packet scheduling with a single queue
AU - Yu, Zhuolong
AU - Hu, Chuheng
AU - Wu, Jingfeng
AU - Sun, Xiao
AU - Braverman, Vladimir
AU - Chowdhury, Mosharaf
AU - Liu, Zhenhua
AU - Jin, Xin
N1 - Publisher Copyright:
© 2021 ACM.
PY - 2021/8/9
Y1 - 2021/8/9
N2 - Programmable packet scheduling enables scheduling algorithms to be programmed into the data plane without changing the hardware. Existing proposals either have no hardware implementations for switch ASICs or require multiple strict-priority queues. We present Admission-In First-Out (AIFO) queues, a new solution for programmable packet scheduling that uses only a {single} first-in first-out queue. AIFO is motivated by the confluence of two recent trends: {shallow} buffers in switches and {fast-converging} congestion control in end hosts, that together leads to a simple observation: the decisive factor in a flow's completion time (FCT) in modern datacenter networks is often {which} packets are enqueued or dropped, not the {ordering} they leave the switch. The core idea of AIFO is to maintain a sliding window to track the ranks of recent packets and compute the relative rank of an arriving packet in the window for admission control. Theoretically, we prove that AIFO provides bounded performance to Push-In First-Out (PIFO). Empirically, we fully implement AIFO and evaluate AIFO with a range of real workloads, demonstrating AIFO closely approximates PIFO. Importantly, unlike PIFO, AIFO can run at line rate on existing hardware and use minimal switch resources - -as few as a single queue.
AB - Programmable packet scheduling enables scheduling algorithms to be programmed into the data plane without changing the hardware. Existing proposals either have no hardware implementations for switch ASICs or require multiple strict-priority queues. We present Admission-In First-Out (AIFO) queues, a new solution for programmable packet scheduling that uses only a {single} first-in first-out queue. AIFO is motivated by the confluence of two recent trends: {shallow} buffers in switches and {fast-converging} congestion control in end hosts, that together leads to a simple observation: the decisive factor in a flow's completion time (FCT) in modern datacenter networks is often {which} packets are enqueued or dropped, not the {ordering} they leave the switch. The core idea of AIFO is to maintain a sliding window to track the ranks of recent packets and compute the relative rank of an arriving packet in the window for admission control. Theoretically, we prove that AIFO provides bounded performance to Push-In First-Out (PIFO). Empirically, we fully implement AIFO and evaluate AIFO with a range of real workloads, demonstrating AIFO closely approximates PIFO. Importantly, unlike PIFO, AIFO can run at line rate on existing hardware and use minimal switch resources - -as few as a single queue.
KW - data center networks
KW - in-network processing
KW - packet scheduling
KW - programmable networks
UR - https://www.scopus.com/pages/publications/85113239584
U2 - 10.1145/3452296.3472887
DO - 10.1145/3452296.3472887
M3 - Conference contribution
AN - SCOPUS:85113239584
T3 - SIGCOMM 2021 - Proceedings of the ACM SIGCOMM 2021 Conference
SP - 179
EP - 193
BT - SIGCOMM 2021 - Proceedings of the ACM SIGCOMM 2021 Conference
PB - Association for Computing Machinery, Inc
T2 - 2021 Annual Conference of the ACM Special Interest Group on Data Communication on the Applications, SIGCOMM 2021
Y2 - 23 August 2021 through 27 August 2021
ER -