Occamy: A Preemptive Buffer Management for On-chip Shared-memory Switches

TL;DR

The paper tackles the mismatch between shrinking on-chip buffers and growing datacenter switch capacity by introducing Occamy, a preemptive buffer management approach that uses redundant memory bandwidth to actively expel packets from over-allocated queues. It decouples admission from expulsion, employs a DT-derived admission policy with a small proactive reserve, and performs reactive, round-robin expulsion of over-allocated queues to reclaim buffer quickly. The authors present a hardware-cost analysis, a P4-based prototype on Intel Tofino, a DPDK software prototype, and large-scale simulations, showing up to 57% better burst absorption and up to 55% improvements in query completion times, along with strong performance isolation under mixed workloads. Taken together, Occamy demonstrates the viability and practical impact of preemptive BM on modern on-chip switch architectures, and it lays groundwork for further exploration and potential industry adoption.

Abstract

Today's high-speed switches employ an on-chip shared packet buffer. The buffer is becoming increasingly insufficient as it cannot scale with the growing switching capacity. Nonetheless, the buffer needs to face highly intense bursts and meet stringent performance requirements for datacenter applications. This imposes rigorous demand on the Buffer Management (BM) scheme, which dynamically allocates the buffer across queues. However, the de facto BM scheme, designed over two decades ago, is ill-suited to meet the requirements of today's network. In this paper, we argue that shallow-buffer switches, intense bursts, along with dynamic traffic call for a highly agile BM that can quickly adjust the buffer allocation as traffic changes. However, the agility of the current BM is fundamentally limited by its non-preemptive nature. Nonetheless, we find that preemptive BM, considered unrealizable in history, is now feasible on modern switch chips. We propose Occamy, a preemptive BM that can quickly adjust buffer allocation. Occamy utilizes the redundant memory bandwidth to actively reclaim and reallocate the over-allocated buffer. Testbed experiments and large-scale simulations show that Occamy can improve the end-to-end performance by up to ~55%.

Figures (23)

• Figure 1: Structure of on-chip shared-memory switch
• Figure 2: Structure of packet buffer
• Figure 3: Dynamic behavior of DT
• Figure 4: An 8-input Maximum Finder based on binary comparator tree
• Figure 5: The buffer choking problem