Serverless computing is increasingly used for data-processing applications in both science and business domains. At the core of serverless data-processing systems is the scheduler, which ensures dynamic decisions about task and data placement. Due to the variety of user, cluster, and workload properties, the design space for high-performance and cost-effective scheduling architectures and mechanisms is vast. The large design space is difficult to explore and characterize. To help the system designer disentangle this complexity, we present ExDe, a framework to systematically explore the design space of scheduling architectures and mechanisms. The framework includes a conceptual model and a simulator to assist in design space exploration. We use the framework, and real-world workloads, to characterize the performance of three scheduling architectures and two mechanisms. Our framework is open-source software available on Zenodo.
%0 Journal Article
%1 TALLURI2023
%A Talluri, Sacheendra
%A Herbst, Nikolas
%A Abad, Cristina
%A De Matteis, Tiziano
%A Iosup, Alexandru
%D 2023
%J Future Generation Computer Systems
%K cloud descartes t_journalmagazine myown
%T ExDe: Design space exploration of scheduler architectures and mechanisms for serverless data-processing
%U https://www.sciencedirect.com/science/article/pii/S0167739X23004211
%X Serverless computing is increasingly used for data-processing applications in both science and business domains. At the core of serverless data-processing systems is the scheduler, which ensures dynamic decisions about task and data placement. Due to the variety of user, cluster, and workload properties, the design space for high-performance and cost-effective scheduling architectures and mechanisms is vast. The large design space is difficult to explore and characterize. To help the system designer disentangle this complexity, we present ExDe, a framework to systematically explore the design space of scheduling architectures and mechanisms. The framework includes a conceptual model and a simulator to assist in design space exploration. We use the framework, and real-world workloads, to characterize the performance of three scheduling architectures and two mechanisms. Our framework is open-source software available on Zenodo.
@article{TALLURI2023,
abstract = {Serverless computing is increasingly used for data-processing applications in both science and business domains. At the core of serverless data-processing systems is the scheduler, which ensures dynamic decisions about task and data placement. Due to the variety of user, cluster, and workload properties, the design space for high-performance and cost-effective scheduling architectures and mechanisms is vast. The large design space is difficult to explore and characterize. To help the system designer disentangle this complexity, we present ExDe, a framework to systematically explore the design space of scheduling architectures and mechanisms. The framework includes a conceptual model and a simulator to assist in design space exploration. We use the framework, and real-world workloads, to characterize the performance of three scheduling architectures and two mechanisms. Our framework is open-source software available on Zenodo.},
added-at = {2023-11-21T14:18:49.000+0100},
author = {Talluri, Sacheendra and Herbst, Nikolas and Abad, Cristina and {De Matteis}, Tiziano and Iosup, Alexandru},
biburl = {https://www.bibsonomy.org/bibtex/2f4ab6d89f485f423e8ec570fb158a0a8/nikolas.herbst},
interhash = {64f7681c27ed8ccf50d41c61e9644c45},
intrahash = {f4ab6d89f485f423e8ec570fb158a0a8},
journal = {Future Generation Computer Systems},
keywords = {cloud descartes t_journalmagazine myown},
note = {https://www.sciencedirect.com/science/article/pii/S0167739X23004211},
timestamp = {2023-11-21T14:18:49.000+0100},
title = {ExDe: Design space exploration of scheduler architectures and mechanisms for serverless data-processing},
url = {https://www.sciencedirect.com/science/article/pii/S0167739X23004211},
year = 2023
}