Autran Macedo

The mechanics of memory-related software aging

Software aging is a phenomenon defined as the continuing degradation of software systems during runtime, being particularly noticeable in long-running applications. Memory-related aging effects are one of the most important problems in this research field. Therefore understanding their causes and how they work is a major requirement in designing dependable software systems. In this paper we go deep into how memory management works inside application process, focusing on two memory problems that cause software aging: fragmenting and leakage. We explain the mechanics of memory-related software aging effects dissecting a real and widely adopted memory allocator. Along with the theoretical explanation, we present an experimental study that illustrates how memory fragmenting and leakage occur and how they accumulate over time in order to cause system aging-related failures.

Monitoring Memory-Related Software Aging: An Exploratory Study

The accumulating effects of software aging have direct influence to the rate of aging-related failures. So far, the most investigated software aging effects are memory related, such as memory leak and memory fragmentation problems. In this work, we present a practical body of knowledge to support the solid understanding of the most important issues in monitoring memory-related software aging effects, focusing on memory leak problems. We discuss important drawbacks of using well known system-wide and application-specific aging indicators, as well as propose effective solutions for both cases.

A Systematic Mapping Review of the First 20 Years of Software Aging and Rejuvenation Research

Although software aging and rejuvenation (SAR) is a young research field, in its first 20 years a lot of knowledge has been produced. Nowadays, important scientific journals and conferences include SAR-related topics in their scope of interest. This fast growing and wide range of dissemination venues pose a challenge to researchers to keep tracking of the new findings and trends in this area. In this work, we collected and analyzed SAR research data using analytics to detect trends, patterns, and thematic gaps, in order to provide a comprehensive view of this research field over its first 20 years. We adopted the systematic mapping approach to answer research questions such as: How the main topics investigated in SAR have evolved over time? Which are the most investigated aging effects? Which rejuvenation techniques and strategies are more frequently used?

Performance Analysis of Control Charts Techniques Applied to IP Traffic Forecasts

Research advances in network traffic forecasting are continually emerging in the literature. Although the progress in this field, the quality control of traffic forecasts is commonly neglected in previous works. In this paper we analyze the performance of CuSum, Shewhart, and EWMA control charts applied to the residuals of forecasting models fitted to selected traffic samples obtained from real IP networks. Based on numerical and graphical analyses, we verify that CuSum and Shewhart control charts perform worse than EWMA, since they report false alarms for several forecasting processes considered under control. Differently, the EWMA control chart shows efficacy to monitor the different types of studied forecast residuals. Finally, we present a sensitive analysis of the EWMA model with respect to the lambda parameter in order to evaluate its flexibility on different traffic scenarios.

Balancing Quality and Performance of Task Distribution in Workflow Based on Resource Aptitude

The distribution of tasks in a workflow system, in general, does not consider resource aptitude. Resources that share the same role in a workflow can have different degrees of aptitude to a task. This work proposes an extension to task distribution strategies and presents the use of this extension in the selection of resources to tasks. The experiments presented here demonstrate the effectiveness of the extension. A comparison with traditional strategies has shown gains in terms of time and quality of tasks executed.

An experimental comparison analysis of kernel-level memory allocators

Memory management is one of the most important tasks of an operating system, since it has significant impact on the whole computing systems performance. The kernel memory allocator is the main OS subsystem responsible for this task, which has to deal with problems intrinsic to dynamic memory allocations, such as memory fragmentation and allocation overhead. In this paper, we present an experimental comparison analysis of four real implementations of kernel memory allocators: SLAB, SLOB, SLUB, and SLQB. This study considered the execution time, memory consumption, and memory fragmentation degree of all allocators under the same workload. Based on the findings obtained experimentally, in most of the cases, the memory allocator with best general performance was the SLOB, followed by SLAB, SLUB, and SLQB.

Evaluation of Compound System Calls in the Linux Kernel on Physical and Virtual Machines

The high use of system calls in many applications has motivated research works focusing on reducing the overhead caused by these calls. In this work we implement three types of compound system calls, which are evaluated experimentally by taking into account their execution time in systems with 1 to 8 cores. We also investigate the system calls running in both physical and virtual machines. All proposed compound system calls present performance gains when compared to their conventional counterparts. The results indicate reduction in the execution time up to 58 percent. The dataset obtained through controlled experiments is analyzed using ANOVA and Tukey tests. We also verify that the highest improvements occur in the non-virtualized environment.

A Comparison of Memory Allocators for Multicore and Multithread Applications: A Quantitative Approach

The performance of memory allocation operations is a very important aspect to be considered in software design, however it is frequently neglected. This paper presents a comparative analysis of seven largely adopted memory allocators. Unlike other related works, based on artificial benchmark tests, we evaluate the selected allocators using real-world middleware applications. In order to compare the performance of the investigated allocators we consider the response time, memory consumption, and memory fragmentation. All tests are evaluated with respect to different combinations of processor cores. The results indicate that for workloads based on memory allocations up to 64 bytes and all combinations of processor cores up to four, the best average response time and memory usage is obtained using the TCMalloc memory allocator, followed by the Ptmalloc version 3.

Evaluation of Compound System Calls in the Linux Kernel

The overhead caused by system calls in many applications has motivated research works focusing on reducing their execution costs. In this work we implement three types of compound system calls, which are evaluated experimentally by taking into account their execution time in systems with one up to eight cores. We also investigate the system calls running in both physical and virtual machines. All proposed compound system calls present performance gains when compared to their conventional counterparts. The results indicate reduction in the execution time up to 58 percent. The dataset obtained through controlled experiments is analyzed using ANOVA and Tukey tests. We also verify that the highest improvements occur in the non-virtualized environment.