Memcached Design on High Performance RDMA Capable Interconnects

Jose, Jithin; Subramoni, Hari; Luo, Miao; Zhang, Minjia; Huang, Jian; Wasi-ur-Rahman, Md.; Islam, Nusrat Sharmin; Ouyang, Xiangyong; Wang, Hao; Sur, Sayantan; Panda, Dhabaleswar K.

doi:10.1109/icpp.2011.37

Cited by 138 publications

(40 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…RDMA has been used in key-value stores, and these systems exhibit good performance and scalability [14,24,26]. On 20 machines, FaRM [14] can serve 150M lookups/second.…”

Section: Related Workmentioning

confidence: 99%

Distributed join algorithms on thousands of cores

et al. 2017

View full text Add to dashboard Cite

Traditional database operators such as joins are relevant not only in the context of database engines but also as a building block in many computational and machine learning algorithms. With the advent of big data, there is an increasing demand for efficient join algorithms that can scale with the input data size and the available hardware resources.In this paper, we explore the implementation of distributed join algorithms in systems with several thousand cores connected by a low-latency network as used in high performance computing systems or data centers. We compare radix hash join to sort-merge join algorithms and discuss their implementation at this scale. In the paper, we explain how to use MPI to implement joins, show the impact and advantages of RDMA, discuss the importance of network scheduling, and study the relative performance of sorting vs. hashing. The experimental results show that the algorithms we present scale well with the number of cores, reaching a throughput of 48.7 billion input tuples per second on 4,096 cores.

show abstract

“…RDMA has been used in key-value stores, and these systems exhibit good performance and scalability [14,24,26]. On 20 machines, FaRM [14] can serve 150M lookups/second.…”

Section: Related Workmentioning

confidence: 99%

Distributed join algorithms on thousands of cores

et al. 2017

View full text Add to dashboard Cite

show abstract

“…Wang and colleagues have extended Hadoop to take advantage of InfiniBand [3]. Similarly in spirit, works by Jose et al [4] and by the authors themselves [5] have investigated how a distributed memory cache can exploit HPC interconnects such as InfiniBand or those of the Blue Gene/P supercomputer for better efficiency. LibRIPC strives to generalize such efforts both with respect to the cloud workloads running atop and the hardware architectures running underneath.…”

Section: Related Workmentioning

confidence: 98%

Light-weight remote communication for high-performance cloud networks

Kehne

Hillenbrand

Stoess

et al. 2012

2012 IEEE 1st International Conference on Cloud Networking (CLOUDNET)

View full text Add to dashboard Cite

Abstract-In this paper, we present early experiences with libRIPC, a light-weight communication library for highperformance cloud networks. Coming cloud networks are expected to be tightly interconnected and to show capabilities formerly reserved to high-performance computing. LibRIPC aims to bring the benefits of such architectures to heterogeneous cloud workloads.LibRIPC was designed for low footprint and easy integration; it supports reconfiguration and mutually untrusted communication partners. LibRIPC offers short and long transmit primitives, which are optimized for control messages and bulk data transfer respectively. Early experiments with a Java-based web server indicate that libRIPC integrates well into typical cloud workloads and brings substantial speedup of at least a factor of three for larger data transfers compared to socket-based TCP/IP communication.

show abstract

“…In addition to kernel bypass, some low-latency object stores rely on RDMA to offload protocol processing on dedicated Infiniband host channel adapters [Dragojevic et al 2014;Jose et al 2011;Mitchell et al 2013;Ousterhout et al 2015]. RDMA can reduce latency but requires that specialized adapters be present at both ends of the connection.…”

Section: Alternative Approachesmentioning

confidence: 99%

“…Application bugs and crashes can corrupt the networking stack and impact other workloads. Other systems go a step further by also replacing TCP/IP with RDMA in order to offload network processing to specialized adapters [Dragojevic et al 2014;Jose et al 2011;Mitchell et al 2013;Ousterhout et al 2015]. However, such adapters must be present at both ends of the connection and can only be used within the datacenter.…”

Section: Introductionmentioning

confidence: 99%

The IX Operating System

Belay¹,

Prekas

Primorac

et al. 2016

ACM Trans. Comput. Syst.

View full text Add to dashboard Cite

The conventional wisdom is that aggressive networking requirements, such as high packet rates for small messages and μs-scale tail latency, are best addressed outside the kernel, in a user-level networking stack. We present IX, a dataplane operating system that provides high I/O performance and high resource efficiency while maintaining the protection and isolation benefits of existing kernels.IX uses hardware virtualization to separate management and scheduling functions of the kernel (control plane) from network processing (dataplane). The dataplane architecture builds upon a native, zero-copy API and optimizes for both bandwidth and latency by dedicating hardware threads and networking queues to dataplane instances, processing bounded batches of packets to completion, and eliminating coherence traffic and multicore synchronization. The control plane dynamically adjusts core allocations and voltage/frequency settings to meet service-level objectives.We demonstrate that IX outperforms Linux and a user-space network stack significantly in both throughput and end-to-end latency. Moreover, IX improves the throughput of a widely deployed, key-value store by up to 6.4× and reduces tail latency by more than 2×. With three varying load patterns, the control plane saves 46%-54% of processor energy, and it allows background jobs to run at 35%-47% of their standalone throughput.

show abstract

Memcached Design on High Performance RDMA Capable Interconnects

Cited by 138 publications

References 7 publications

Distributed join algorithms on thousands of cores

Distributed join algorithms on thousands of cores

Light-weight remote communication for high-performance cloud networks

The IX Operating System

Contact Info

Product

Resources

About