Distributed Queue-Based Locking Using Advanced Network Features

Devulapalli, A.; Wyckoff, P.

doi:10.1109/icpp.2005.34

Cited by 13 publications

(11 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our results shown in Figure 5 present the basic performance improvement that our scheme (N-CoShED) shows over existing schemes: (i) basic Distributed Queue based Non-shared Locking (DQNL) [10] and (ii) traditional Send/Receive-based Server Locking (SRSL). N-CoShED scheme shows 39% improvement over the SRSL scheme.…”

Section: Distributed Lock Managermentioning

confidence: 76%

Designing Efficient Systems Services and Primitives for Next-Generation Data-Centers

Vaidyanathan

Narravula

Balaji

et al. 2007

2007 IEEE International Parallel and Distributed Processing Symposium

View full text Add to dashboard Cite

Current data-centers lack in efficient support for intelligent services, such as requirements for caching documents and cooperation of caching servers, efficiently monitoring and managing the limited physical resources, load-balancing, controlling overload scenarios, that are becoming a common requirement today. On the other hand, the System Area Network (SAN) technology is making rapid advances during the recent years. Besides high performance, these modern interconnects are providing a range of novel features and their support in hardware (e.g., RDMA, atomic operations). In this paper, we extend our previously proposed framework comprising of three layers (communication protocol support, data-center service primitives and advanced data-center services) that work together to tackle the issues associated with existing data-centers. We present the performance results using data-center services such as cooperative caching and active resource monitoring and data-center primitives such as distributed data sharing substrate and distributed lock manager, which demonstrate significant performance benefits achievable by our framework as compared to existing data-centers in several cases.

show abstract

Section: Distributed Lock Managermentioning

confidence: 76%

Designing Efficient Systems Services and Primitives for Next-Generation Data-Centers

Vaidyanathan

Narravula

Balaji

et al. 2007

2007 IEEE International Parallel and Distributed Processing Symposium

View full text Add to dashboard Cite

show abstract

“…[2,5] presented a detailed analysis of lock management in distributed systems. There are a few works which tried to leverage one-sided RDMA verbs to build a DLM [3,6]. The authors in [3] proposed to use atomic operations to provide.…”

Section: Methodsmentioning

confidence: 99%

“…There are a few works which tried to leverage one-sided RDMA verbs to build a DLM [3,6]. The authors in [3] proposed to use atomic operations to provide. The basic idea is to have the server only store the tail of the FIFO lists.…”

Section: Methodsmentioning

confidence: 99%

Using RDMA for Lock Management

Chung,

Zamanian

2015

Preprint

View full text Add to dashboard Cite

In this work, we aim to evaluate different Distributed Lock Management service designs with Remote Direct Memory Access (RDMA). In specific, we implement and evaluate the centralized and the RDMAenabled lock manager designs for fast network settings. Experimental results confirms a couple of hypotheses. First, in the traditional centralized lock manager design, CPU is the bottleneck and bypassing CPU on client-to-server communication using RDMA results in better lock service perofrmance. Second, different lock manager designs with RDMA in consideration result in even better performance; we need to re-design lock management system for RDMA and fast networks.

show abstract

“…Researchers in [5] have proposed distributed queue based DLM using RDMA operations. Though this work exploits the benefits of RDMA operations for locking services, their design can only support exclusive mode locking.…”

Section: Related Workmentioning

confidence: 99%

Designing passive synchronization for MPI-2 one-sided communication to maximize overlap

Santhanaraman

Narravula

Panda

2008

2008 IEEE International Symposium on Parallel and Distributed Processing

View full text Add to dashboard Cite

Scientific computing has seen an immense growth in recent years. MPI has become the de facto standard for parallel programming model for distributed memory systems. MPI-2 standard expanded MPI to include onesided communications. Computation and communication overlap is an important goal for one-sided applications. While the passive synchronization mechanism for MPI-2 one-sided communication allows for good overlap, the actual overlap achieved is often limited by the design of both the MPI library and the application.In this paper we aim to improve the performance of MPI-2 one-sided communication. In particular, we focus on the following important aspects: (i) designing one-sided passive synchronization (Direct Passive) support using InfiniBand atomic operations to handle both exclusive as well as shared locks (ii) enhancing one-sided communication progress to provide scope for better overlap that one-sided applications can leverage. (iii) study the overlap potential of passive synchronization and its impact on applications. We demonstrate the possible benefits of our approaches for the MPI-2 SPLASH LU application benchmark. Our results show an improvement of up to 87% for a 64 process run over the existing design.

show abstract

Distributed Queue-Based Locking Using Advanced Network Features

Cited by 13 publications

References 18 publications

Designing Efficient Systems Services and Primitives for Next-Generation Data-Centers

Designing Efficient Systems Services and Primitives for Next-Generation Data-Centers

Using RDMA for Lock Management

Designing passive synchronization for MPI-2 one-sided communication to maximize overlap

Contact Info

Product

Resources

About