Abstract
Recent achievements in high-performance computing significantly narrow the performance gap between single and multi-node computing, and open up opportunities for systems with remote shared memory. The combination of in-memory storage, remote direct memory access and remote calls requires rethinking how data organized, protected and queried in distributed systems. Reviewed models let us implement new interpretations of distributed algorithms allowing us to validate different approaches to avoid race conditions, decrease resource acquisition or synchronization time. In this paper, we describe the data model for mixed memory access with analysis of optimized data structures. We also provide the result of experiments, which contain a performance comparison of data structures, operating with different approaches, evaluate the limitations of these models, and show that the model does not always meet expectations. The purpose of this paper to assist developers in designing data structures that will help to achieve architectural benefits or improve the design of existing distributed system.
Highlights
Distributed computing is used in a wide variety of fields
Software part consist of operating system (OS) SUSE Linux Enterprise Server 15, we used GCC 10.1 as compiler and C++17, STL, Berkeley Container Library (BCL), BOOST as libraries/components
We introduced the access model to shared data structures located in virtual global memory space
Summary
Distributed computing is used in a wide variety of fields. This includes scientific research, technical developments such as facial recognition, control systems and autopilots [1,2,3,4]. Many of them will consider this area as the main one for the modernization of business processes and the basic tool for making business decisions in the future This means that each of these companies will need some kind of raw data processing — most likely through distributed clusters. These kind of systems uses data structures which may be hard to implement using traditional message passing interfaces in a distributed memory environment. Conclusion and future work section are used to discuss the results of experiments and possible open problems for future research, such as high performance file systems, security and reliability of RDMA-based solutions
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
