本文關(guān)鍵詞：云計算環(huán)境下低成本存儲科學(xué)數(shù)據(jù)的演化CTT-SP算法　出處：《計算機科學(xué)》2017年12期 　論文類型：期刊論文

  更多相關(guān)文章： 數(shù)據(jù)存儲 計算存儲平衡 云計算 科學(xué)應(yīng)用 成本

【摘要】：云計算系統(tǒng)強大的計算能力和存儲容量,使得科學(xué)家可以在其上部署計算型和數(shù)據(jù)密集型的應(yīng)用,并把大量的應(yīng)用數(shù)據(jù)存儲在云計算環(huán)境下。基于云服務(wù)即用即付模型,針對原有數(shù)據(jù)存儲狀態(tài),考慮云服務(wù)價格變化所產(chǎn)生的狀態(tài)調(diào)整成本,同時為降低存儲大量生成的科學(xué)數(shù)據(jù)的成本,在傳統(tǒng)最小成本基準(zhǔn)的CTT-SP算法的基礎(chǔ)上,提出了一種演化CTT-SP算法。在云計算環(huán)境下針對云服務(wù)的新價格,該算法可自動決定所生成的科學(xué)數(shù)據(jù)是否需要存儲,從而使計算和存儲達(dá)到更佳的平衡。以亞馬遜的成本模型為例,對大量隨機數(shù)據(jù)集進(jìn)行實驗,結(jié)果表明,當(dāng)云服務(wù)價格變化后,所提演化CTT-SP算法有效地降低了存儲科學(xué)數(shù)據(jù)的總成本。
[Abstract]:Cloud computing systems have powerful computing power and storage capacity, allowing scientists to deploy computational and data-intensive applications on them. And a lot of application data is stored in cloud computing environment. Based on the cloud service pay-as-you-go model, the state adjustment cost caused by the change of cloud service price is considered in view of the original data storage state. At the same time, in order to reduce the cost of storing a large number of generated scientific data, based on the traditional minimum cost benchmark CTT-SP algorithm. An evolutionary CTT-SP algorithm is proposed, which can automatically determine whether the generated scientific data needs to be stored for the new price of cloud services in the cloud computing environment. In order to achieve a better balance between computing and storage. Taking Amazon's cost model as an example, a large number of random data sets are tested, the results show that, when the price of cloud services changes. The proposed evolutionary CTT-SP algorithm effectively reduces the total cost of storing scientific data.
【作者單位】： 廣東工業(yè)大學(xué)計算機學(xué)院;悉尼大學(xué)電力信息工程學(xué)院;斯威本科技大學(xué)軟件與電子工程學(xué)院;
【基金】：廣東省產(chǎn)學(xué)研重點項目(2014XYD-007) 廣東省科技計劃項目(2012B091000173)資助
【分類號】：TP333;TP393.09
【正文快照】： 統(tǒng)),按需的云計算服務(wù)給科學(xué)研究部署帶來了便利[4],但科1　引言學(xué)研究者為使用的云資源所支付訴成本仍然很高。特別地,近年來,云計算作為新興的并行分布式計算模式,根據(jù)用科學(xué)應(yīng)用越來越趨向于數(shù)據(jù)密集型的應(yīng)用[5],其產(chǎn)生的數(shù)據(jù)戶需求,提供冗余、廉價和可擴展的資源[1]。云計，

本文編號：1384079