FULL-KV: Flexible and Ultra-Low-Latency In-Memory Key-Value Store System Design on CPU-FPGA
FULL-KV: Flexible and Ultra-Low-Latency In-Memory Key-Value Store System Design on CPU-FPGA
复制标题
DOI：
10.1109/tpds.2020.2973965
复制
发表时间：
2020-08
影响因子：
5.3
通讯作者：
Yunhui Qiu;Jinyu Xie;Hankun Lv;Wenbo Yin;W. Luk;Lingli Wang;Bowei Yu;Hua Chen;X. Ge;Zhijian Liao;X. Shi
中科院分区：
文献类型：
--
作者：
Yunhui Qiu;Jinyu Xie;Hankun Lv;Wenbo Yin;W. Luk;Lingli Wang;Bowei Yu;Hua Chen;X. Ge;Zhijian Liao;X. Shi
内存键值存储（IMKVS）在数据中心中非常流行。然而，大数据带来了巨大的挑战，性能和功耗，因为通用冯诺依曼计算机架构。支持零拷贝网络的远程直接内存访问（RDMA）技术可以部分缓解这个问题，但对于KVS来说仍然效率低下。为了克服这个问题，我们提出了一个灵活的和超低延迟的IMKVS系统命名为FULL-KV，基于CPU-FPGA异构体系结构。FPGA用作KVS加速器，可以绕过CPU，并使用高度并行的硬件架构实现网络堆栈和KVS处理。对于PUT/GET操作，FULL-KV的系统延迟可以低至1.5μs/2.2μs，比当前最先进的基于硬件的KVS系统快3.0倍/1.5倍。此外，FULL-KV可以支持4倍更大的值（高达4 M字节）。给定20 Gbps的总以太网带宽，单节点FULL-KV的峰值吞吐量可以达到每秒2600万次键值操作（Mops）。在采用商用以太网交换机的两节点测试系统中，峰值吞吐量可达52 Mops，体现了系统的可扩展性和实用性。
In-memory key-value store (IMKVS) has gained great popularity in data centers. However, big data brings great challenges in performance and power consumption because of the general-purpose Von Neumann computer architecture. Remote direct memory access (RDMA) technology supporting zero-copy networking could partly alleviate the problem but is still not efficient for KVS. To overcome this problem, we present a flexible and ultra-low-latency IMKVS system named FULL-KV, based on a CPU-FPGA heterogeneous architecture. The FPGA serves as a KVS accelerator that can bypass the CPU and implement both the network stacks and the KVS processing with a highly parallel hardware architecture. The system latency of FULL-KV can achieve as low as 1.5μs/2.2μs for the PUT/GET operation, which is 3.0x/1.5x faster than current state-of-the-art hardware-based KVS systems. Besides, FULL-KV can support 4x larger values (up to 4M bytes). Given a total Ethernet bandwidth of 20Gbps, the peak throughput of the single-node FULL-KV can reach 26.0 million key-value operations per second (Mops). In the two-node test system with a commercial Ethernet switch, the peak throughput can reach 52Mops, manifesting the system scalability and practicability.