CPU demand for web serving: Measurement analysis and dynamic estimation

Managing the resources in a large Web serving system requires knowledge of the resource needs for service requests of various types. In order to investigate the properties of Web traffic and its demand, we collected measurements of throughput and CPU utilization and performed some data analyses. First, we present our findings in relation to the time-varying nature of the traffic, the skewness of traffic intensity among the various types of requests, the correlation among traffic streams, and other system-related phenomena. Then, given such nature of web traffic, we devise and implement an on-line method for the dynamic estimation of CPU demand.

Assessing resource needs is commonly performed using techniques such as off-line profiling, application instrumentation, and kernel-based instrumentation. Little attention has been given to the dynamic estimation of dynamic resource needs, relying only on external and high-level measurements such as overall resource utilization and request rates. We consider the problem of dynamically estimating dynamic CPU demands of multiple kinds of requests using CPU utilization and throughput measurements. We formulate the problem as a multivariate linear regression problem and obtain its basic solution. However, as our measurement data analysis indicates, one is faced with issues such as insignificant flows, collinear flows, space and temporal variations, and background noise. In order to deal with such issues, we present several mechanisms such as data aging, flow rejection, flow combining, noise reduction, and smoothing. We implemented these techniques in a Work Profiler component that we delivered as part of a broader system management product. We present experimental results from using this component in scenarios inspired by real-world usage of that product.     

The Impact of an Agile Methodology on the Well Being of Development Teams

This paper describes an empirical study, which addresses the aspect of well being amongst members of the software development teams. The question of interest is whether an agile methodology has any distinct effect on the well being of the software developers. Both quantitative and qualitative methods were utilised, including the participative observation, focus group interviews, close-ended questionnaires and simple statistical tests such as Spearman Correlation and Mann–Whitney test. Initial results showed that an agile methodology (XP) has a positive effect on the level of enthusiasm of the software developers in the most dynamic project. To understand why XP can increase enthusiasm, results are interpreted with references to cognitive, affective and managerial properties of the practices studied. This result needs further investigation on the individual effects of each practice on the wellbeing and attitudes of Software Engineering (SE) teams.     

Acute cardio-respiratory changes induced by hyperpnoea using a respiratory muscle trainer

The aim of this study was to examine the cardio-respiratory effects of voluntary hyperpnoea using a respiratory muscle trainer (RMT) with three different sized rebreathing bags. In particular, the effects of hyperpnoea on inspired and end-tidal gas concentrations were determined. Seven adult males completed three 30 min bouts of hyperpnoea using optimal, oversized and undersized rebreathing bags. Inspired (F(I)) and expired end-tidal (F(ET)) O2 and CO2 concentrations, arterial O2 saturation (S(AO2)) and heart rate were measured during hyperpnoea. Before and after a bout of hyperpnoea, pulmonary function and blood pressure (BP) were assessed. Data were analysed using a two-way repeated-measures ANOVA, with p < 0.05 considered significant. Three subjects experienced discomfort during hyperpnoea and stopped after 20 min. During hyperpnoea, the F(ETCO2) was maintained at 4.6 +/- 0.7% irrespective of bag size. The increase in F(ICO2) over time reached 0.5 +/- 0.5% at 20 min. The F(IO2) fell to 19.4 +/- 0.8% at 20 min, and S(AO2) decreased to 97%. Heart rate and systolic BP increased slightly, but independently of rebreathing bag volume. No changes in pulmonary function or diastolic BP were found. It is concluded that the RMT maintained a constant F(ETCO2) at the expense of a mild hypoxia. The acute effects of hyperpnoea on the cardio-respiratory system are generally mild, but not always tolerable for 30 min.     

Nonlinear Soil–Abutment–Bridge Structure Interaction for Seismic Performance-Based Design

Current seismic design of bridges is based on a displacement performance philosophy using nonlinear static pushover analysis. This type of bridge design necessitates that the geotechnical engineer predict the resistance of the abutment backfill soils, which is inherently nonlinear with respect to the displacement between soil backfill and the bridge structure. This paper employs limit-equilibrium methods using mobilized logarithmic-spiral failure surfaces coupled with a modified hyperbolic soil stress–strain behavior (LSH model) to estimate abutment nonlinear force-displacement capacity as a function of wall displacement and soil backfill properties. The calculated force-displacement capacity is validated against the results from eight field experiments conducted on various typical structure backfills. Using LSH and experimental data, a simple hyperbolic force-displacement (HFD) equation is developed that can provide the same results using only the backfill soil stiffness and ultimate soil capacity. HFD is compatible with current CALTRANS practice in regard to the seismic design of bridge abutments. The LSH and HFD models are powerful and effective tools for practicing engineers to produce realistic bridge response for performance-based bridge design.     

A computational framework for automating generation of sizing function in assembly meshing via disconnected skeletons

This paper proposes a framework for generating mesh sizing functions for assembly models. Size control is crucial in obtaining a high-quality mesh with a reduced number of elements. The reduction in the number of elements will decrease computation time and memory use during mesh generation and analysis. The framework consists of a background octree lattice for storing the sizing function, a set of source entities for providing sizing information based on geometric information, and an interpolation module for calculating the sizing on the background octree lattice using the source entities. Source entities are generated by performing a detailed systematic study to identify all the geometric factors of an assembly. Disconnected skeletons are extracted and used as tools to measure 3D proximity and 2D proximity, which are two of the geometric factors. The framework facilitates the generation of a variety of meshes with a low computational cost, to meet industry needs. The framework has been tested on many industrial parts, and sizing control on a few typical assemblies has been presented to demonstrate the effectiveness of the proposed framework.     

A network model of barrier synchronization algorithms

We consider two algorithms for the barrier synchronization ofN processes: the Dissemination algorithm⁽²⁾ and Brooks algorithm.^(1,2) Both algorithms comprise a number of binary communications amongst the processes, organized into a sequence of stages. It is shown that Brooks' algorithm⁽¹⁾ requires between LogN(log₂ N) and 2 LogN stages, the lower bound being guaranteed only in the case thatN is a power of 2 (cubic) and the upper bound seemingly needed for most otherN. On the other hand, it is shown⁽²⁾ that the Dissemination algorithm requires only LogN stages regardless ofN, making it apparently superior to the Brooks algorithm. We introduce a network model of local barrier algorithms. Using it we obtain a rigorous correctness proof for local barrier algorithms, and show that the number of stages in the Brooks algorithm is bounded above by LogN+1. The Brooks algorithms is therefore essentially equivalent in time complexity to the Dissemination algorithm. We then address the question of which values ofN admit exactly LogN Brooks stages. We find a sufficient condition, and conjecture that it is also necessary.     

Ion-based quantum simulation of many-body electron–electron Coulomb interaction

We propose an exact mathematical mapping that can be useful for making an analog quantum simulator that uses ion-based systems to realize the many-body electron–electron Coulomb interaction of an electron gas. This exact mathematical mapping allows us to deal with a system that is difficult to solve and control using a potentially more experimentally feasible setup. We show that ions can efficiently simulate electronic Coulomb interactions by using a unitary dilatation transform. The transformation does not need to be physically implemented if only the energy spectrum is desired, eliminating the complexity overhead. This proposal works in any number of dimensions and could be used to simulate different topological phases of electrons in graphene-like structures, by using ions confined in honeycomb lattices.