Dynamic resource allocation for shared data centers using online measurements
Dynamic resource allocation for shared data centers using online measurements
复制标题
DOI：
10.1007/3-540-44884-5_21
复制
发表时间：
2003-06
期刊：
影响因子：
--
通讯作者：
A. Chandra;W. Gong;Prashant J. Shenoy
中科院分区：
文献类型：
--
作者：
A. Chandra;W. Gong;Prashant J. Shenoy
万维网的日益普及导致了托管第三方网络应用程序和服务的互联网数据中心的出现。在这样的数据中心中，应用程序所有者租用服务器资源，作为回报，应用程序获得了资源可用性和性能的保证。为了提供这样的保证，数据中心必须提供足够的资源来满足应用程序的需求。由于已知web工作负载会随时间动态变化，动态资源分配技术对于在共享数据中心上运行的web应用程序提供保证是必要的。为了解决这个问题，我们使用了一个将在线测量与预测和资源分配技术相结合的系统架构。为了捕获应用程序工作负载的瞬态行为，我们使用广义处理器共享（GPS）服务器的时域描述对服务器资源进行建模。该模型将应用程序资源需求与其动态变化的工作负载特征联系起来。该模型的参数使用在线监测和预测框架不断更新。该框架使用时间序列分析技术，根据测量的系统度量来预测预期的工作负载参数。然后，我们采用约束非线性优化技术，根据估计的应用程序需求动态分配服务器资源。我们的技术的主要优点是，它们捕获了应用程序的瞬态行为，同时将非线性纳入系统模型，而不像基于稳态系统行为[6]或线性系统模型[1]的技术。
The growing popularity of the World Wide Web has led to the advent of Internet data centers that host third-party web applications and services. In such data centers, the application owner rents server resources, and in return, the application is provided guarantees on resource availability and performance. To provide such guarantees, the data center must provision sufficient resources to meet application needs. Since web workloads are known to vary dynamically with time, dynamic resource allocation techniques are necessary to provide guarantees to web applications running on shared data centers. To address this issue, we use a system architecture that combines online measurements with prediction and resource allocation techniques. To capture the transient behavior of the application workloads, we model a server resource using a time-domain description of a generalized processor sharing (GPS) server. This model relates application resource requirements to their dynamically changing workload characteristics. The parameters of this model are continuously updated using an online monitoring and prediction framework. This framework uses time series analysis techniques to predict expected workload parameters from measured system metrics. We then employ a constrained non-linear optimization technique to dynamically allocate the server resources based on the estimated application requirements. The main advantage of our techniques is that they capture the transient behavior of applications while incorporating nonlinearity in the system model unlike techniques based on steady-state system behavior [6] or linear system models [1].