Integrating Adaptive Components: An Emerging Challenge in Performance-Adaptive Systems and a Server Farm Case-Study
Integrating Adaptive Components: An Emerging Challenge in Performance-Adaptive Systems and a Server Farm Case-Study
复制标题
DOI：
10.1109/rtss.2007.48
复制
发表时间：
2007-12
期刊：
影响因子：
--
通讯作者：
Jin Heo;D. Henriksson;Xue Liu;T. Abdelzaher
中科院分区：
文献类型：
--
作者：
Jin Heo;D. Henriksson;Xue Liu;T. Abdelzaher
性能敏感软件系统的复杂性增加导致自动适应策略的使用增加，而不是手动性能调整。然而，自适应成分成分为更大的自适应系统，提出了各自适应性政策中潜在不兼容引起的挑战。因此，不稳定或调整较差的反馈回路可能导致性能恶化。本文（i）提出了一种称为适应图分析的机制，用于识别组成的适应性策略之间的潜在不兼容，并且（ii）说明了一种共同适应的一般设计方法，可以解决这种不相容性。我们的结果是通过在多层网络服务器农场中受到软实时限制的多层网络服务器农场中的能量最小化的案例研究来证明的。两项独立有效的节能策略（在不需要时开/关策略，将机器关闭，动态电压缩放策略）显示出冲突，导致合并后的能源消耗增加。我们的适应图分析预测了问题，我们的共同适应设计方法可以找到提高性能的解决方案。经营行业标准TPC-W电子商务基准的17个服务器农场的实验结果表明，当工作量不高的同时，在可接受的范围内保持潜伏期，同时将能源消耗降低超过50％ 。该论文是拟议的冲突识别和解决方法的概念证明，并邀请进一步研究科学以组成自适应系统。
The increased complexity of performance-sensitive software systems leads to increased use of automated adaptation policies in lieu of manual performance tuning. Composition of adaptive components into larger adaptive systems, however, presents challenges that arise from potential incompatibilities among the respective adaptation policies. Consequently, unstable or poorly-tuned feedback loops may result that cause performance deterioration. This paper (i) presents a mechanism, called adaptation graph analysis, for identifying potential incompatibilities between composed adaptation policies and (ii) illustrates a general design methodology for co-adaptation that resolves such incompatibilities. Our results are demonstrated by a case study on energy minimization in multi-tier Web server farms subject to soft real-time constraints. Two independently efficient energy saving policies (an on/off policy that switches machines off when not needed and a dynamic voltage scaling policy) are shown to conflict leading to increased energy consumption when combined. Our adaptation graph analysis predicts the problem, and our co-adaptation design methodology finds a solution that improves performance. Experimental results from a 17-server farm running the industry standard TPC-W e-commerce benchmark show that co-adaptation renders a cut-down in energy consumption by more than 50%, when workload is not high, while maintaining latency within acceptable bounds. The paper serves as a proof of concept of the proposed conflict-identification and resolution methodology and an invitation to further investigate a science for composing adaptive systems.