A Catalogue of Green Architectural Tactics for the Cloud
Status:: 🟩
Links:: 30_Knowledge/Green Architectural Tactics
Metadata
Authors:: Procaccianti, Giuseppe; Lago, Patricia; Lewis, Grace A.
Title:: A Catalogue of Green Architectural Tactics for the Cloud
Date:: 2014
URL:: http://ieeexplore.ieee.org/document/6976608/
DOI:: 10.1109/MESOCA.2014.12
Bibliography
Procaccianti, G., Lago, P., & Lewis, G. A. (2014). A Catalogue of Green Architectural Tactics for the Cloud. 2014 IEEE 8th International Symposium on the Maintenance and Evolution of Service-Oriented and Cloud-Based Systems, 29–36. https://doi.org/10.1109/MESOCA.2014.12
Zotero
Type:: #zotero/conferencePaper
Zotero::
Keywords:: [✅, Architectural Tactics, Cloud computing, Cloud Computing, Energy consumption, Energy Efficiency, Green Software, Monitoring, Software Architecture]
Relations
Related:: @Vos.etal.2022.ArchitecturalTacticsOptimize
Related:: @Jagroep.etal.2017.ExtendingSoftwareArchitecture
Related:: @Paradis.etal.2021.ArchitecturalTacticsEnergy
Related:: @Procaccianti.etal.2014.GreenArchitecturalTactics
Abstract
Energy efficiency is a primary concern for the ICT sector. In particular, the widespread adoption of cloud computing technologies has drawn attention to the massive energy consumption of data centers. Although hardware constantly improves with respect to energy efficiency, this should also be a main concern for software. In previous work we analyzed the literature and elicited a set of techniques for addressing energy efficiency in cloud-based software architectures. In this work we codified these techniques in the form of Green Architectural Tactics. These tactics will help architects extend their design reasoning towards energy efficiency and to apply reusable solutions for greener software.
Notes & Annotations
🟨 Note (last modified: 2024-01-13#14:33:06) (
Extends the paper “Green Architectural Tactics for the Cloud”
📑 Annotations (imported on 2024-01-13#16:22:25)
A recent study [2] showed that migrating all business applications in the U.S. to the cloud could reduce their energy footprint by 87%.
[2] E. Masanet, A. Shehabi, L. Ramakrishnan, J. Liang, X. Ma, B. Walker, V. Hendrix, and P. Maantha, “The Energy Efficiency Potential of Cloudbased Software: A U.S. Case Study,” Laurence Berkeley National Lab, California, Tech. Rep., June 2013.
This work extends a paper published in [7], where we presented three scenarios for energy efficiency and an example tactic for each scenario. In this paper we extend our initial work, by presenting our view of energy efficiency as a quality attribute and a full catalogue of Green Architectural Tactics with examples.
[7] G. Procaccianti, P. Lago, and G. Lewis, “Green architectural tactics for the cloud,” in Proceedings of the 11th Working IEEE/IFIP Conference on Software Architecture (WICSA), Apr. 2014, to appear.
Some preliminary investigations go back to the work of Rangaraj & Bahsoon [9], who used market-based economics theory to define a framework for optimizing power consumption in energy-unaware software architectures at runtime. Bahsoon then planned to apply the same approach to cloud architectures [10].
[9] G. Rangaraj and R. Bahsoon, “Green software architectures: A marketbased approach,” in The Second International Workshop on Software Research and Climate Change (WSRCC), 2010.
[10] R. Bahsoon, “A framework for dynamic self-optimization of power and dependability requirements in green cloud architectures,” in Proceedings of the 4th European conference on Software architecture (ECSA’10). Berlin, Heidelberg: Springer-Verlag, 2010, pp. 510–514.
In [11], Seo et al. come closer to our objective by defining a framework that estimates the energy consumption of three distributed system architectural styles. Their goal is to evaluate the most appropriate architectural style before implementation.
[11] C. Seo, G. Edwards, D. Popescu, S. Malek, and N. Medvidovic, “A framework for estimating the energy consumption induced by a distributed system’s architectural style,” in Proceedings of the 8th international workshop on Specification and verification of component based systems (SAVCBS ’09). New York, NY, USA: ACM, 2009, pp. 27–34.
@Seo.etal.2009.FrameworkEstimatingEnergy
Cyber-foraging, a form of mobile cloud computing in which mobile devices offload expensive computation to more powerful servers in the cloud, is a common strategy for saving battery power on mobile devices [14]. However, it is not uncommon for literature on cyber-foraging to refer to the cloud as having infinite resources; which means that no reusable cyber-foraging strategies have been defined yet for architecting energy-aware software systems that address the EE of the system as a whole.
[14] M. Satyanarayanan, “Pervasive computing: vision and challenges,” Personal Communications, IEEE, vol. 8, no. 4, pp. 10–17, 2001.
According to Bass et al. [15], EE is to be regarded as a “system” quality attribute because it is the result of an indirect action of software. However, Bass et al. also argue that the line between “software” and “system” quality attributes is very thin. In the end, even if energy is ultimately consumed by hardware, it is software that determines hardware behavior.
[15] L. Bass, P. Clements, and R. Kazman, Software architecture in practice, 3rd ed. Addison-Wesley, 2012.
4th edition: @Bass.etal.2021.SoftwareArchitecturePractice
In previous work [5] we identified a set of recurring design solutions, described in the literature, to achieve EE in cloudbased software architectures. In this work, we codified these solutions as tactics – that is, “design decisions that influence the achievement of a quality attribute response” [15].
[5] G. Procaccianti, S. Bevini, and P. Lago, “Energy efficiency in cloud software architectures,” in Proceedings of the 27th Conference on Environmental Informatics (ENVIROINFO), vol. 1. Shaker Verlag GmbH, 2013, pp. 291–299.
the Consolidation tactic concentrates the VM instances on the minimum number of servers needed. Powering down the unused servers will evidently increase the EE of the cloudbased software.
Elasticity has, of course, a direct connection with EE: the more closely resource provisioning matches demand, the more energy efficient the infrastructure is. The Scaling Down tactic allows to adapt resource provisioning, while the Workload Scheduling tactic is meant to prioritize and assign the load to the different virtual resources in order to match the demand.