Josep Lluis Berral

Energy-Aware Scheduling in Virtualized Datacenters

The reduction of energy consumption in large-scale datacenters is being accomplished through an extensive use of virtualization, which enables the consolidation of multiple workloads in a smaller number of machines. Nevertheless, virtualization also incurs some additional overheads (e.g. virtual machine creation and migration) that can influence what is the best consolidated configuration, and thus, they must be taken into account. In this paper, we present a dynamic job scheduling policy for power-aware resource allocation in a virtualized datacenter. Our policy tries to consolidate workloads from separate machines into a smaller number of nodes, while fulfilling the amount of hardware resources needed to preserve the quality of service of each job. This allows turning off the spare servers, thus reducing the overall datacenter power consumption. As a novelty, this policy incorporates all the virtualization overheads in the decision process. In addition, our policy is prepared to consider other important parameters for a datacenter, such as reliability or dynamic SLA enforcement, in a synergistic way with power consumption. The introduced policy is evaluated comparing it against common policies in a simulated environment that accurately models HPC jobs execution in a virtualized datacenter including power consumption modeling and obtains a power consumption reduction of 15% with respect to typical policies.

Adaptive on-line software aging prediction based on machine learning

The growing complexity of software systems is resulting in an increasing number of software faults. According to the literature, software faults are becoming one of the main sources of unplanned system outages, and have an important impact on company benefits and image. For this reason, a lot of techniques (such as clustering, fail-over techniques, or server redundancy) have been proposed to avoid software failures, and yet they still happen. Many software failures are those due to the software aging phenomena. In this work, we present a detailed evaluation of our chosen machine learning prediction algorithm (M5P) in front of dynamic and non-deterministic software aging. We have tested our prediction model on a three-tier web J2EE application achieving acceptable prediction accuracy against complex scenarios with small training data sets. Furthermore, we have found an interesting approach to help to determine the root cause failure: The model generated by machine learning algorithms.

Web Customer Modeling for Automated Session Prioritization on High Traffic Sites

In the Web environment, user identification is becoming a major challenge for admission control systems on high traffic sites. When a web server is overloaded there is a significant loss of throughput when we compare finished sessions and the number of responses per second; longer sessions are usually the ones ending in sales but also the most sensitive to load failures. Session-based admission control systems maintain a high QoS for a limited number of sessions, but does not maximize revenue as it treats all non-logged sessions the same. We present a novel method for learning to assign priorities to sessions according to the revenue that will generate. For this, we use traditional machine learning techniques and Markov-chain models. We are able to train a system to estimate the probability of the users purchasing intentions according to its early navigation clicks and other static information. The predictions can be used by admission control systems to prioritize sessions or deny them if no resources are available, thus improving sales throughput per unit of time for a given infrastructure. We test our approach on access logs obtained from a high-traffic online travel agency, with promising results.

Adaptive distributed mechanism against flooding network attacks based on machine learning

Adaptive techniques based on machine learning and data mining are gaining relevance in self-management and self-defense for networks and distributed systems. In this paper, we focus on early detection and stopping of distributed flooding attacks and network abuses. We extend the framework proposed by Zhang and Parashar (2006) to cooperatively detect and react to abnormal behaviors before the target machine collapses and network performance degrades. In this framework, nodes in an intermediate network share information about their local traffic observations, improving their global traffic perspective. In our proposal, we add to each node the ability of learning independently, therefore reacting differently according to its situation in the network and local traffic conditions. In particular, this frees the administrator from having to guess and manually set the parameters distinguishing attacks from non-attacks: now such thresholds are learned and set from experience or past data. We expect that our framework provides a faster detection and more accuracy in front of distributed flooding attacks than if static filters or single-machine adaptive mechanisms are used. We show simulations where indeed we observe a high rate of stopped attacks with minimum disturbance to the legitimate users.

Tailoring Resources: The Energy Efficient Consolidation Strategy Goes Beyond Virtualization

Virtualization and consolidation are two complementary techniques widely adopted in a global strategy to reduce system management complexity. In this paper we show how two simple and well-known techniques can be combined to dramatically increase the energy efficiency of a virtualized and consolidated data center. This result is obtained by introducing a new approach to the consolidation strategy that allows an important reduction in the amount of active nodes required to process a web workload without degrading the offered service level. Furthermore, when the system eventually gets overloaded and no energy can be saved without loosing performance, we show how these techniques can still improve the overall value obtained from the workload. The two techniques are memory compression and request discrimination, and were separately studied and validated in a previous work to be now combined in a joint effort. Our results indicate that an important improvement can be achieved by deciding not only how resources are allocated, but also how they are used. Moreover, we believe that this serves as an illustrative example of a new way of management: tailoring the resources to meet high level energy efficiency goals.

ALOJA-ML: A Framework for Automating Characterization and Knowledge Discovery in Hadoop Deployments

This article presents ALOJA-Machine Learning (ALOJA-ML) an extension to the ALOJA project that uses machine learning techniques to interpret Hadoop benchmark performance data and performance tuning; here we detail the approach, efficacy of the model and initial results. The ALOJA-ML project is the latest phase of a long-term collaboration between BSC and Microsoft, to automate the characterization of cost-effectiveness on Big Data deployments, focusing on Hadoop. Hadoop presents a complex execution environment, where costs and performance depends on a large number of software (SW) configurations and on multiple hardware (HW) deployment choices. Recently the ALOJA project presented an open, vendor-neutral repository, featuring over 16.000 Hadoop executions. These results are accompanied by a test bed and tools to deploy and evaluate the cost-effectiveness of the different hardware configurations, parameter tunings, and Cloud services. Despite early success within ALOJA from expert-guided benchmarking, it became clear that a genuinely comprehensive study requires automation of modeling procedures to allow a systematic analysis of large and resource-constrained search spaces. ALOJA-ML provides such an automated system allowing knowledge discovery by modeling Hadoop executions from observed benchmarks across a broad set of configuration parameters. The resulting empirically-derived performance models can be used to forecast execution behavior of various workloads; they allow a-priori prediction of the execution times for new configurations and HW choices and they offer a route to model-based anomaly detection. In addition, these models can guide the benchmarking exploration efficiently, by automatically prioritizing candidate future benchmark tests. Insights from ALOJA-MLs models can be used to reduce the operational time on clusters, speed-up the data acquisition and knowledge discovery process, and importantly, reduce running costs. In addition to learning from the methodology presented in this work, the community can benefit in general from ALOJA data-sets, framework, and derived insights to improve the design and deployment of Big Data applications.

Power-Aware Multi-data Center Management Using Machine Learning

The cloud relies upon multi-data center (multi-DC) infrastructures distributed along the world, where people and enterprises pay for resources to offer their web-services to worldwide clients. Intelligent management is required to automate and manage these infrastructures, as the amount of resources and data to manage exceeds the capacities of human operators. Also, it must take into account the cost of running the resources (energy) and the quality of service towards web-services and clients. (De-)consolidation and priming proximity to clients become two main strategies to allocate resources and properly place these web-services in the multi-DC network. Here we present a mathematical model to describe the scheduling problem given web-services and hosts across a multi-DC system, enhancing the decision makers with models for the system behavior obtained using machine learning. After running the system on real DC infrastructures we see that the model drives web-services to the best locations given quality of service, energy consumption, and client proximity, also (de-)consolidating according to the resources required for each web-service given its load.

Empowering automatic data-center management with machine learning

The Cloud as computing paradigm has become nowadays crucial for most Internet business models. Managing and optimizing its performance on a moment-by-moment basis is not easy given as the amount and diversity of elements involved (hardware, applications, workloads, customer needs...). Here we show how a combination of scheduling algorithms and data mining techniques helps improving the performance and profitability of a data-center running virtualized web-services. We model the data-centers main resources (CPU, memory, IO), quality of service (viewed as response time), and workloads (incoming streams of requests) from past executions. We show how these models to help scheduling algorithms make better decisions about job and resource allocation, aiming for a balance between throughput, quality of service, and power consumption.

From performance profiling to predictive analytics while evaluating hadoop cost-efficiency in ALOJA

During the past years the exponential growth of data, its generation speed, and its expected consumption rate presents one of the most important challenges in IT both for industry and research. For these reasons, the ALOJA research project was created by BSC and Microsoft as an open initiative to increase cost-efficiency and the general understanding of Big Data systems via automation and learning. The development of the project over its first year, has resulted in a open source benchmarking platform used to produce the largest public repository of Big Data results1, featuring over 42,000 job execution details. ALOJA also includes web-based analytic tools to evaluate and gather insights about cost-performance of benchmarked systems. The tools offer means to extract knowledge that can lead to optimize configuration and deployment options in the Cloud i.e., selecting the most cost-effective VMs and cluster sizes. This article describes the evolution of the project focus and research lines, for a period of over a year while continuously benchmarking systems for Big Data. As well discusses the motivation - both technical and market-based - of such changes. It also presents the main results from the evaluation of different OS and Hadoop configurations, covering over 100 hardware deployments. During this time, ALOJAs initial target has shifted from a previous low-level profiling of Hadoop runtime with HPC tools, passing through extensive benchmarking and evaluation of a large body of results via aggregation, to currently leveraging Predictive Analytics (PA) techniques. The ongoing efforts in PA show promising results to automatically model the behavior of systems i.e., predicting job execution times with high accuracy or to reduce the number of benchmark runs needed. As well as for Knowledge Discovery (KD) to find relations among software and hardware components. Techniques that jointly support foresighting cost-effectiveness of new defined systems, reducing benchmarking time and costs.

J2EE instrumentation for software aging root cause application component determination with AspectJ

Unplanned system outages have a negative impact on company revenues and image. While the last decades have seen a lot of efforts from industry and academia to avoid them, they still happen and their impact is increasing. According to many studies, one of the most important causes of these outages is software aging. Software aging phenomena refers to the accumulation of errors, usually provoking resource contention, during long running application executions, like web applications, which normally cause applications/systems hang or crash. Determining the software aging root cause failure, not the resource or resources involved in, is a huge task due to the growing day by day complexity of the systems. In this paper we present a monitoring framework based on Aspect Programming to monitor the resources used by every application component in runtime. Knowing the resources used by every component of the application we can determine which components are related to the software aging. Furthermore, we present a case study where we evaluate our approach to determine in a web application scenario, which components are involved in the software aging with promising results.