Vanish Talwar

No "power" struggles: coordinated multi-level power management for the data center

Power delivery, electricity consumption, and heat management are becoming key challenges in data center environments. Several past solutions have individually evaluated different techniques to address separate aspects of this problem, in hardware and software, and at local and global levels. Unfortunately, there has been no corresponding work on coordinating all these solutions. In the absence of such coordination, these solutions are likely to interfere with one another, in unpredictable (and potentially dangerous) ways. This paper seeks to address this problem. We make two key contributions. First, we propose and validate a power management solution that coordinates different individual approaches. Using simulations based on 180 server traces from nine different real-world enterprises, we demonstrate the correctness, stability, and efficiency advantages of our solution. Second, using our unified architecture as the base, we perform a detailed quantitative sensitivity analysis and draw conclusions about the impact of different architectures, implementations, workloads, and system design choices.

GViM: GPU-accelerated virtual machines

The use of virtualization to abstract underlying hardware can aid in sharing such resources and in efficiently managing their use by high performance applications. Unfortunately, virtualization also prevents efficient access to accelerators, such as Graphics Processing Units (GPUs), that have become critical components in the design and architecture of HPC systems. Supporting General Purpose computing on GPUs (GPGPU) with accelerators from different vendors presents significant challenges due to proprietary programming models, heterogeneity, and the need to share accelerator resources between different Virtual Machines (VMs). To address this problem, this paper presents GViM, a system designed for virtualizing and managing the resources of a general purpose system accelerated by graphics processors. Using the NVIDIA GPU as an example, we discuss how such accelerators can be virtualized without additional hardware support and describe the basic extensions needed for resource management. Our evaluation with a Xen-based implementation of GViM demonstrate efficiency and flexibility in system usage coupled with only small performance penalties for the virtualized vs. non-virtualized solutions.

vManage: loosely coupled platform and virtualization management in data centers

Management is an important challenge for future enterprises. Previous work has addressed platform management (e.g., power and thermal management) separately from virtualization management (e.g., virtual machine (VM) provisioning, application performance). Coordinating the actions taken by these different management layers is important and beneficial, for reasons of performance, stability, and efficiency. Such coordination, in addition to working well with existing multi-vendor solutions, also needs to be extensible to support future new management solutions potentially operating on different sensors and actuators. In response to these requirements, this paper proposes vManage, a solution to loosely couple platform and virtualization management and facilitate coordination between them in data centers. Our solution is comprised of registry and proxy mechanisms that provide unified monitoring and actuation across platform and virtualization domains, and coordinators that provide policy execution for better VM placement and runtime management, including a formal approach to ensure system stability from inefficient management actions. The solution is instantiated in a Xen environment through a platform-aware virtualization manager at a cluster management node, and a virtualization-aware platform manager on each server. Experimental evaluations using enterprise benchmarks show that compared to traditional solutions, vManage can achieve additional power savings (10% lower power) with significantly improved service-level guarantees (71% less violations) and stability (54% fewer VM migrations), at low overhead.

Power Management of Datacenter Workloads Using Per-Core Power Gating

While modern processors offer a wide spectrum of software-controlled power modes, most datacenters only rely on dynamic voltage and frequency scaling (DVFS, a.k.a. P-states) to achieve energy efficiency. This paper argues that, in the case of datacenter workloads, DVFS is not the only option for processor power management. We make the case for per-core power gating (PCPG) as an additional power management knob for multi-core processors. PCPG is the ability to cut the voltage supply to selected cores, thus reducing to almost zero the leakage power for the gated cores. Using a testbed based on a commercial 4-core chip and a set of real-world application traces from enterprise environments, we have evaluated the potential of PCPG. We show that PCPG can significantly reduce a processors energy consumption (up to 40%) without significant performance overheads. When compared to DVFS, PCPG is highly effective saving up to 30% more energy than DVFS. When DVFS and PCPG operate together they can save up to almost 60%.

Using Asymmetric Single-ISA CMPs to Save Energy on Operating Systems

CPUs consume too much power. Modern complex cores sometimes waste power on functions that are not useful for the code they run. In particular, operating system kernels do not benefit from many power-consuming features intended to improve application performance. We advocate asymmetric single-ISA multicore systems, in which some cores are optimized to run OS code at greatly improved energy efficiency.

Statistical techniques for online anomaly detection in data centers

Online anomaly detection is an important step in data center management, requiring light-weight techniques that provide sufficient accuracy for subsequent diagnosis and management actions. This paper presents statistical techniques based on the Tukey and Relative Entropy statistics, and applies them to data collected from a production environment and to data captured from a testbed for multi-tier web applications running on server class machines. The proposed techniques are lightweight and improve over standard Gaussian assumptions in terms of performance.

Online detection of utility cloud anomalies using metric distributions

The online detection of anomalies is a vital element of operations in data centers and in utility clouds like Amazon EC2. Given ever-increasing data center sizes coupled with the complexities of systems software, applications, and workload patterns, such anomaly detection must operate automatically, at runtime, and without the need for prior knowledge about normal or anomalous behaviors. Further, detection should function for different levels of abstraction like hardware and software, and for the multiple metrics used in cloud computing systems. This paper proposes EbAT - Entropy-based Anomaly Testing - offering novel methods that detect anomalies by analyzing for arbitrary metrics their distributions rather than individual metric thresholds. Entropy is used as a measurement that captures the degree of dispersal or concentration of such distributions, aggregating raw metric data across the cloud stack to form entropy time series. For scalability, such time series can then be combined hierarchically and across multiple cloud subsystems. Experimental results on utility cloud scenarios demonstrate the viability of the approach. EbAT outperforms threshold-based methods with on average 57.4% improvement in accuracy of anomaly detection and also does better by 59.3% on average in false alarm rate with a ‘near-optimum’ threshold-based method.

A flexible architecture integrating monitoring and analytics for managing large-scale data centers

To effectively manage large-scale data centers and utility clouds, operators must understand current system and application behaviors. This requires continuous, real-time monitoring along with on-line analysis of the data captured by the monitoring system, i.e., integrated monitoring and analytics -- Monalytics [28]. A key challenge with such integration is to balance the costs incurred and associated delays, against the benefits attained from identifying and reacting to, in a timely fashion, undesirable or non-performing system states. This paper presents a novel, flexible architecture for Monalytics in which such trade-offs are easily made by dynamically constructing software overlays called Distributed Computation Graphs (DCGs) to implement desired analytics functions. The prototype of Monalytics implementing this flexible architecture is evaluated with motivating use cases in small scale data center experiments, and a series of analytical models is used to understand the above trade-offs at large scales. Results show that the approach provides the flexibility needed to meet the demands of autonomic management at large scale with considerably better performance/cost than traditional and brute force solutions.

Cloud Management: Challenges and Opportunities

Cloud computing offers on-demand access to shared resources and services, hosted in warehouse sized data centers at cloud providers. Effective management of these shared resources and services is one of the key requirements for the delivery of cloud computing. However, there are several challenges to achieve effective cloud management. These include scale, multiple levels of abstraction, federation, sustainability, and dynamism. In this paper, we outline these challenges, and then describe specific examples of new management architectures that address these challenges. We focus on driving principles for the new designs, and give an illustration of its deployment on Open Cirrus - a research cloud test bed. Overall, the paper provides open issues and challenges in cloud management for the research community to address.

Comparison of Approaches to Service Deployment

IT today is driven by the trend of increasing scale and complexity. Utility and grid computing models, PlanetLab, and traditional data centers, are reaching the scale of thousands of computers. Installed software consists of dozens of interdependent applications and services. As the complexity and scale of these systems continues to grow, it becomes increasingly difficult to administer and manage them. At the same time, the service deployment technologies are still based on scripts and configuration files with minimal ability to express dependencies, to document and to verify configurations. This results in hard-to-use and erroneous system configurations. Language- and model-based tools, such as SmartFrog and Radio, are proposed for addressing these deployment challenges, but it is unclear whether they are beneficial over traditional solutions. In this paper, we quantitatively compare manual, script-, language-, and model-based deployment solutions as a function of scale, complexity, and susceptibility to change. We also qualitatively compare them in terms of expressiveness and barrier to first use. We demonstrate that script-based solutions are well matched for large scale deployments, language-based for services of large complexity, and model-based for dynamic changes to the design. Finally, we offer a table summarizing rules of thumb regarding which solution to use in which case, subject to deployment needs