Aviad Sela

Autonomic self-optimization according to business objectives

A central challenge in the runtime management of computing environments is the necessity to keep these environments continuously optimized. In this paper we introduce a new paradigm, which focuses on self-optimization according to high-level business objectives such as maximizing revenues. It replaces the more traditional optimizations that are based upon IT measures such as resource availability. A general, autonomous process is defined to enable such optimizations, and a set of technologies and methodologies is introduced to support the implementation of such a process. The paper concludes with two types of validation tests carried out on an eCommerce site, that demonstrate the value and applicability of this approach.

Pluggable Analysis Viewpoints for Design Space Exploration

Abstract Viewpoint modeling is an effective approach for analyzing and designing complex systems. Splitting various elements and corresponding constraints into different perspectives of interests, enables separation of concerns such as domains of expertise, levels of abstraction, and stages in lifecycle. Specifically, in Systems Engineering different viewpoints could include functional requirements, physical architecture, safety, geometry, timing, scenarios, etc. Despite partial interdependences, the models are usually developed independently by different parties, using different tools and languages. However, the essence of Systems Engineering requires repetitive integration of many viewpoints in order to find feasible designs and to make good architectural decisions, e.g., in each mapping between consecutive levels of abstraction and in each design space exploration. This integration into one consistent model becomes a significant challenge from both modeling and information management perspectives. In this paper we suggest (1) a unique modular algebraic viewpoint representation robust to design evolution and suitable for generation of the integrated optimization/analysis models, and (2) an underlying ontology-based approach for consistent integration of local viewpoint concepts into the unified design space model. We show an example of an optimization model with different combinations of partially interdependent Analysis Viewpoints. Using the proposed modeling and information management approaches the underlying viewpoints equations can be applied without modification, making the approach pluggable.

A scalable heterogeneous solution for massive data collection and database loading

Massive collection of data at high rates is critical for many industries. Typically, a massive stream of records is gathered from the business information network at a very high rate. Because of the complexity of the collection process, the classical database solution falls short. The high volume and rate of records involved requires a heterogeneous pipeline comprised of two major parts: a system that carries out massive collection and then uploads the information to a database, and a subsequent data analysis and management system consisting of an Extract Transform and Load component. We developed a massive collection and loading system, based on a highly scalable heterogeneous architecture solution. The solution has been applied successfully for Telco revenue assurance, and can be applied to other industrial areas. The solution was successful in scaling up a Telco client system to handle streams of records ten times larger than was previously possible.

A methodological framework for business-oriented modeling of IT infrastructure

The creation of IT simulation models for uses such as capacity planning and optimization is becoming more and more widespread. Traditionally, the creation of such models required deep modeling and/or programming expertise, thus severely limiting their extensive use. Moreover, many modern intelligent tools now require simulation models in order to carry out their function. For these tools to be widely deployable, the derivation of simulation models must be made possible without requiring excessive technical knowledge. Hence we introduce a general methodology that enables an almost automatic deployment of IT simulation models, based on three fundamental principles: Modeling only at the required level of detail; modeling standard components using pre-prepared models; and automatically deriving the application-specific model details. The technical details underlying this approach are presented. In addition, a case study, showing the application of this methodology to an eCommerce site, demonstrates the applicability of this approach.

Reusable Derivation of Operational Metrics for Architectural Optimization

Abstract Maintaining coherence between system functional, performance, production and operational requirements is a key to the ability to optimize the design of large-scale systems. Different architectural configurations entail significant differences in functionality, performance, ease of manufacturing/assembly and operational behavior. While the first two are the usual concerns in architectural tradeoff analysis, the last two, reflected by manufacturability and operational metrics, such as manufacturability and affordability, are often neglected in architectural optimization. In this work, we propose a methodology to derive the formal specification of operational metrics applicable to design optimization based on life cycle processes, such as “manufacturing”, “sunny day operation”, and “unplanned maintenance”. These operational metrics are presented in the context of an industrial case study for an Unmanned Underwater Vehicle (UUV) to provide context for the recommended approach. We suggest an approach to (1) define libraries of reusable operational metrics based on architectural properties, (2) build reusable data processing patterns to calculate these architectural properties, and (3) map calculated architectural parameters to a specific design model.