dist-gem5: Distributed simulation of computer clusters

Alian, Mohammad; Darbaz, Umur; Dózsa, Gábor; Diestelhorst, Stephan; Kim, Daehoon; Kim, Nam Sung

doi:10.1109/ispass.2017.7975287

Cited by 28 publications

(10 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, the simulation of both systems is executed within the same thread; thus, a serious performance penalty is triggered while no synchronization primitives between the two systems are provided. (Recently, authors in Mohammad et al [2017] provide a distributed GEM5 version that supports only ARM-based processors and a simplistic network model. )…”

Section: Supporting Parallel/distributedmentioning

confidence: 99%

“…both for academic/research and commercial purposes makes the design and deployment of such systems and applications a very interesting and expanding area. In the HPC domain, there are certain tools such as those by Mohammad et al [2017], which model processing units in a cycle accurate way, while others (e.g., see Ahmed et al [2017]) that focus on the simulation of real networks; thus, the existing approaches either focus on providing high accurate results per node and more simplistic network models or they replace processing cycle-accurate simulation with application traces or functional modeling and focus on the network part. COSSIM goes beyond those approaches by combining successfully both aspects while providing additional functionality (such as testing the robustness of applications).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Novel, Highly Integrated Simulator for Parallel and Distributed Systems

Tampouratzis

Papaefstathiou

Nikitakis

et al. 2020

ACM Trans. Archit. Code Optim.

View full text Add to dashboard Cite

In an era of complex networked parallel heterogeneous systems, simulating independently only parts, components, or attributes of a system-under-design is a cumbersome, inaccurate, and inefficient approach. Moreover, by considering each part of a system in an isolated manner, and due to the numerous and highly complicated interactions between the different components, the system optimization capabilities are severely limited. The presented fully-distributed simulation framework (called as COSSIM) is the first known open-source, highperformance simulator that can handle holistically system-of-systems including processors, peripherals and Extension of Conference Paper: N. Tampouratzis et al. An Open-Source Extendable, Highly-Accurate, and Security Aware CPS Simulator. 2017 13th International Conference on DCOSS. When we presented COSSIM at DCOSS, it was a CPStailored simulator with limited capabilities and the corresponding evaluation results were, thus, also limited. COSSIM can now support the simulation of Cloud and HPC infrastructures as well, so the work presented in this article extends the work in DCOSS in several aspects: (a) we extended the synchronization system in a novel way so as to allow it to be decentralized/distributed itself while we evaluated it utilizing multiple HLA servers on different physical machines to demonstrate that COSSIM is a fully-distributed framework, with no centralized components. In the conference paper, a simple centralized approach was described, and (b) we extend the COSSIM capabilities in order to support a wide range of different real network protocols by developing the micro-router functionality. In the conference paper, COSSIM could support only wired Ethernet interconnections, so (c) we extended its capabilities so as to allow for different tradeoff between accuracy and simulation times; no such capabilities have been supported in the 1st version of COSSIM. (d) We have extended our complete framework so as to be able to handle a thousand simulated nodes; the original COSSIM simulator could support up to 32 processing nodes. (e) Several real-world use cases of parallel applications are ported to our new system and evaluated across different metrics. For example, we evaluated the new version of COSSIM when simulating up to 1024 X86-based and ARM-based nodes in a cluster that contains 200 physical cores. In addition, in the current article, COSSIM has been evaluated in terms of all the additional features implemented (i.e., accuracy, scalability, and performance) when executing the widely-used Netperf benchmark suite, which allows it to be compared against similar solutions that may appear in the future. Furthermore, a new real-world use case has been added in the evaluation section (Section 6.4), which clearly demonstrates that COSSIM accurately simulates a complete real system. (f) Finally, in order to further increase the impact of our work to the HPC/Cloud/CPS community, the complete source code with its documentation has now become freely distributed through GitHub. N. T...

show abstract

Section: Supporting Parallel/distributedmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A Novel, Highly Integrated Simulator for Parallel and Distributed Systems

Tampouratzis

Papaefstathiou

Nikitakis

et al. 2020

ACM Trans. Archit. Code Optim.

View full text Add to dashboard Cite

show abstract

“…Dist-GEM5 tried to simulate several nodes using multiple simulation systems. This simulator uses TCP sockets as a channel for transfer of synchronization and data messages between a switch node and a full-system node, which enables to prevent data messages from avoiding synchronization messages (due to the strict ordering between TCP packets) [8].…”

Section: Dist-gem5mentioning

confidence: 99%

“…This simulator improves the checkpointing mechanism of its previous work pd-GEM5 and is strongly coupled with the Ethernet protocol. Dist-GEM5 can deliver a fast, scalable and detailed infrastructure of simulation for modelling and evaluating large computing groups [8].…”

Section: Dist-gem5mentioning

confidence: 99%

“…In regard parallel system simulation, Mohammed Alian et al [8] have proposed Dist-GEM5 simulation tool to simulate the architectural structure and network behaviour of parallel and distributed heterogeneous computing systems. Dist-GEM5 in [8] creates communication channels for GEM5s installed on standalone devices to communicate via the host network system and to address network simulation tool's (GEM5's) scalability problem over parallel systems. COSSIM [9] can run distributed and parallel heterogeneous system models by integrating cGEM5 with OMNET++.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Appropriate Synchronization Time Allocation for Distributed Heterogeneous Parallel Computing Systems

2019

KSII TIIS

View full text Add to dashboard Cite

Parallel computing system components should be harmonized, and this harmonization is kept existent using synchronization time. Synchronization time affects the system in two ways. First, if we have too little synchronization time, some tasks face the problem of harmonization, as they need appropriate time to update and synchronize with the system. Second, if we allocate a large amount of time, stall system created. Random allocation of synchronization time for parallel systems slows down not only the booting time of the system but also the execution time of each application involved in the system. This paper presents a simulator used to test and allocate appropriate synchronization time for distributed and parallel heterogeneous systems. The simulator creates the parallel and heterogeneous system to be evaluated, and lets the user vary the synchronization time to optimize the booting time. NS3-cGEM5 simulator in this paper is formed by HLA-RTI federation integration of the two independent architecture and network simulators-NS3 and cGEM5. Therefore, nodes created on these simulators need synchronizations for harmonized system performance. We tested and allocated the appropriate synchronization time for our sample parallel system composed of one x86 server and three ARM clients.

show abstract