Karthikeyan Ponnalagu

SOMA-ME: a platform for the model-driven design of SOA solutions

The service-oriented modeling and architecture modeling environment (SOMA-ME) is first a framework for the model-driven design of service-oriented architecture (SOA) solutions using the service-oriented modeling and architecture (SOMA) method. In SOMA-ME, Unified Modeling Language (UML™) profiles extend the UML 2.0 metamodel to domain-specific concepts. SOMA-ME is also a tool that extends the IBM Rational® Software Architect product to provide a development environment and automation features for designing SOA solutions in a systematic and model-driven fashion. Extensibility, traceability, variation-oriented design, and automatic generation of technical documentation and code artifacts are shown to be some of the properties of the SOMA-ME tool.

Measuring the Quality of Service Oriented Design

Service Oriented Architecture (SOA) has gained popularity as a design paradigm for realizing enterprise software systems through abstract units of functionality called services. While the key design principles of SOA have been discussed at length in the literature, much of the work is prescriptive in nature and do not explain how adherence to these principles can be quantitatively measured in practice. In some cases, metrics for a limited subset of SOA quality attributes have been proposed, but many of these measures have not been empirically validated on real-life SOA designs. In this paper, we take a deeper look at how the key SOA quality attributes of service cohesion, coupling, reusability, composability and granularity may be evaluated, based only on service design level information. We survey related work, adapt some of the well-known software design metrics to the SOA context and propose new measures where needed. These measures adhere to mathematical properties that characterize the quality attributes. We study their applicability on two real-life SOA design models from the insurance industry using a metrics computation tool integrated with an Eclipse-based service design environment. We believe that availability of these measures during SOA design will aid early detection of design flaws, allow different design options and trade-offs to be considered and support planning for development, testing and governance of the services.

Morpheus: Semantics-based Incremental Change Propagation in SOA-based Solutions

SOA-based solutions are typically modeled as business processes composed of loosely coupled services. Such an approach promises the flexibility to more easily customize the solution functionality in line with constantly changing customer requirements. The research issue that we address in this paper, therefore, is how to best accomplish this customization. Current approaches are typically manual and rather ad-hoc, involving repeated attempts to synchronize between (higher-level) design artifacts and (lower-level) source code to determine the configuration points. Alternatively, we propose a framework, Morpheus, which uses design artifacts to locate points of change in an SOA based solution via multi-level change propagation. First, we formally define the structure and semantics of the design artifacts, and the relationships among them. Second, we use these relationships to enumerate all possible changes in each design artifact; if two design artifacts share a relationship,then we also map a change in one design artifact to related changes in the other. Third, using these change relationships,we present an algorithm to traverse the design artifacts so as to propagate changes based on change requirements, ultimately resulting in the modifications needed to support the changed requirements. We illustrate our ideas on a simple yet realistic example in the insurance domain, and also present a prototype implementation.

Goal-Driven business process derivation

Solutions to the problem of deriving business processes from goals are critical in addressing a variety of challenges facing the services and business process management community, and in particular, the challenge of quickly generating large numbers of effective process designs (often a bottleneck in industry-scale deployment of BPM). The problem is similar to the planning problem that has been extensively studied in the artificial intelligence (AI) community. However, the direct application of AI planning techniques places an onerous burden on the analyst, and has proven to be difficult in practice. We propose a practical yet rigorous (semi-automated) algorithm for business process derivation from goals. Our approach relies on being able to decompose process goals to a more refined collection of sub-goals whose ontology is aligned with that of the effects of available tasks which can be used to construct the business process. Once process goals are refined to this level, we are able to generate a process design using a procedure that leverages our earlier work on semantic effect annotation of process designs. We illustrate our ideas throughout this paper with a real-life running example, and also present a proof-of-concept prototype implementation.

Model Driven Provisioning in Multi-tenant Clouds

In multi-tenant cloud systems today, provisioning of resources for new tenancy is based on selection from a catalogue published by the cloud provider. The published images are generally a stack of appliances with Infrastructure (IaaS) and Platform (PaaS) layers and optionally Application layers (SaaS). Such a ready-made model enables quicker and streamlined resource provisioning to clients. However, this approach poses certain challenges to clients in the short run and providers in the long run. Unique tenancy requirements from each client are forcibly generalized by selecting one of the available images from the catalogue as the tenancy requirements are not modeled or validated to start with. Moreover, resource provisioning is mostly done towards addressing the peak load expectations in the tenancy. Such a static approach does not help in adapting to dynamically changing tenancy requirements, most often leading to the tenants owning and subsequently paying for more than what they need. In particular, provisioned resources are expected to perform at the same level of quality without accounting for their changing health. In our paper, we propose an extensible dynamic provisioning framework to address these challenges. We start with defining a Tenancy Requirements Model (TRM) which helps map provisioned resources with tenants. The provisioned and candidate resources are also modeled with their Quality of Service (QoS) characteristics which we call Health Grading Model (HGM); this helps in continuous monitoring and grading of resources based on health parameters and enables health prediction for future provisioning. Together, TRM and HGM allow dynamic re-provisioning for existing tenants based on either changing tenancy requirements or health grading predictions. We also present algorithms for prediction based provisioning and tenancy requirement matching. We illustrate our ideas throughout this paper with a running example, and present a proof-of-concept prototype implementation on IBMs Rational Software Architect modeling tool.

Enhancing Asset Search and Retrieval in a Services Repository using Consumption Contexts

Software organizations wanting to implement a systematic reuse program face the challenge of organizing and cataloging their software assets so that they can be retrieved in different contexts of usage across their divisions. With the advent of services-oriented architectures (SOA), of which Web services is an example, software components are readily available as services on the web using standard protocols. However, service descriptions are usually only those that are provided by the service producers and unless a service producer has thought about all the contexts in which its service may be used, there is no guarantee that the service can be retrieved with high recall. In this paper, we investigate how different contexts of asset consumption may be used for better asset modeling and discovery. We introduce an extensible set of consumption factors where the service descriptions may be provided by the service producer or other roles, implement a prototype that can evolve with the modeled factors, and demonstrate that the solution enables improved precision and recall for services available in a large-scale software organization.

Dynamic provisioning in multi-tenant service clouds

Cloud-based systems promise an on-demand service provisioning system along with a “pay-as-you-use” policy. In the case of multi-tenant systems this would mean dynamic creation of a tenant by integrating existing cloud-based services on the fly. Presently, dynamic creation of a tenant is handled by building the required components from scratch. Although multi-tenant systems help providers save cost by allocating multiple tenants to the same instance of an application, they incur huge reconfiguration costs. Cost and time spent on these reconfiguration activities can be reduced by re-constructing tenants from existing tenant configurations supported by service providers. Multi-tenant cloud-based systems also lack the facility of allowing clients to specify their requirements. Giving clients the flexibility to specify requirements helps them avoid spending an excessive amount of time and effort looking through a list of services, many of which might not be relevant to them. Moreover, dynamic provisioning in the cloud requires an integrated solution across the technology stack (software, platform and infrastructure) combining functional, non-functional and resource allocation requirements. Existing research works in the area of web service matching, although numerous, still fall short, since they usually consider each requirement type in isolation and cannot provide an integrated solution. To that end, in this paper we investigate the features needed for dynamic service provisioning on the cloud. We propose a novel User Interface-Tenant Selector-Customizer (UTC) model and approach, which enables cloud-based services to be systematically modeled and provisioned as variants of existing service tenants in the cloud. Our approach considers functional, non-functional and resource allocation requirements, which are explicitly specified by the client via the user interface component of the model. To the best of our knowledge, ours is the first such integrated approach. We illustrate our ideas using a realistic running example, and also present a proof-of-concept prototype built using IBM’s Rational Software Architect modeling tool. We also present experimental results demonstrating the applicability of our matching algorithm. Our results show significant reduction in matching time with the help of an elimination process that reduces the search space needed for performing matching.

Towards a Formal Model for Optimal Task-Site Allocation and Effort Estimation in Global Software Development

Motivated by the desire to cut costs and development effort, software organizations have increasingly adopted a global development approach. However, the cost savings, if any, from this globalization, is often offset by hidden costs such as handoffs between sites, synchronization of development among sites, integration of software developed at distributed sites, language/cultural issues, travel costs, communication costs, etc. Although several empirical studies have been conducted on this issue, due to the lack of an integrated formal model, such studies have not produced consistent and usable results. To that end, in this paper, we present an integrated formal model for analyzing global software development. Our model comprises two parts. First, we consider all tasks in a software project that can be geographically distributed, and the possible sites where they can be allocated. We develop an optimal task-site allocation model. Our approach then generates an effort estimate for the new allocation, which is based on the following factors: expected general percentage allocation of overall effort estimate to each task in the development lifecycle, and effort estimate for executing a task at a particular site (in terms of the effort estimate for executing the same task at the home site, viz., without globalization). The final effort estimate is therefore derived as a function of the effort estimate for executing the overall software project in the home site; this estimate provides project managers with a more accurate understanding of expected cost savings from globalization, if any. Throughout our paper, we illustrate our approach using a real global software development project at IBM as a running example.

Towards a Variability Model for SOA-Based Solutions

The prevalent approach to developing business solutions is driven by requirements expressed as process models. This is essentially a top-down model-driven development approach for building solutions from a given process model specification, from scratch. However, in order to enhance the reusability of such solutions, the emphasis has now shifted to defining business solutions via service-oriented architectures (SOAs) that compose software services to support business processes. This enables creating and reusing services and processes as software “assets” which is more scalable and profitable. In this context, the key to reusing assets is to support the right mechanisms to incrementally refine existing software services as well as business processes. Our earlier works, viz., Variation Oriented Engineering (VOE) and Variation Oriented Service Design (VOSD), proposed some mechanisms in this direction. But our experiences in developing and inducting VOE and VOSD into IBM’s business units motivated the need for a rigorous formal foundation for variability in SOA-based solutions. To that end, this paper proposes a Variability Model for SOA-based solutions, through which variations in a service can be modeled and maintained. Our Variability Model also clarifies the extent to which an existing service can be modified to suit a business process requirement, while ensuring that the overall integrity of the service is maintained. We also present an algorithm for determining whether a variant of a service is a legal variant, and we demonstrate our Variability Model on a realistic running example in the insurance domain, via a prototype built on IBM’s Rational Software Architect modeling tool.

Deriving service variants from business process specifications

Software service organizations typically tend to develop custom solutions upon entry to a project engagement. This is not a scalable proposition. Today, driven by the need for enhancing profitability, reuse of existing assets across customer engagements is a more viable strategy. This is leading to the transformation of software services organizations from a labor-based to an asset-based model. In the enterprise application development domain, business processes are used to model the dynamic behavior of enterprise applications. Hence one highly prevalent asset-based approach in this domain is to build a scalable portfolio of services using a service-oriented architecture (SOA) approach. This provides a mechanism for representing business processes as a set of business-aligned, loosely-coupled services that can be iteratively composed and re-composed to create loosely coupled composite applications that mirror and support business processes. A key research problem that arises here is how to instantiate a stated business process specification as a combination of variants of existing services in the portfolio. We show via an illustrative example that this problem is non-trivial and scalable, and it depends on the semantics of the business process specification, especially with respect to the inter-service data and control flow dependencies that the instantiated business process specification has to satisfy. To that end, we also present our Variation-Oriented Service Design (VOSD) algorithm for automatically deriving service-level variants from the stated business process specifications. We also show that our algorithm is scalable, since its asymptotic complexity is linear in the number of service variants. We also illustrate our algorithm via the example.