Environment

11 pages
6 views

Leveraging Similarities and Defect Analysis Feedback in Migration Program- FSP Approach

of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Description
Leveraging Similarities and Defect Analysis Feedback in Migration Program- FSP Approach
Transcript
   Procedia - Social and Behavioral Sciences 74 ( 2013 ) 408 – 418 1877-0428  © 2013 The Authors. Published by Elsevier Ltd.Selection and/or peer-review under responsibility of IPMA doi: 10.1016/j.sbspro.2013.03.046 26 th  IPMA World Congress, Crete, Greece, 2012 Leveraging Similarities and Defect Analysis Feedback in Migration Programs: FSP Approach Veerendra K. Rai * , Sanjit Mehta, G. Ramanand Systems Research Laboratory, Tata Consultancy Services, 54 B, Hadapsar Industrial Estate, Pune - 411013, India Abstract This paper reports a case study of a large migration program and the benefits of FSP approach to exploit the similarities among applications to be migrated and the defect analysis feedback loop. FSP approach is an acronym for- F  ix the core issue-Fix  S  imilar defects-  P  revent occurrence of similar defects.  The paper argues that there are substantial similarities in large migration programs and these could be exploited to expedite the causal analysis to reduce the time taken in inspection and fault detection and helps prevent faults thus reducing the effort and cost taken in migration. This paper modelled the FSP approach and simulated it with program data. Results validate the claim made by FSP approach with increasing similarity. Gains made by FSP approach in reducing the work with undetected faults and fault prevention are substantial. Effectiveness of FSP approach is dependent on degree of similarity among the applications being migrated. It appears from results FSP is essentially a quality assurance approach and reduction in effort and cost of migration are added benefits. FSP approach involves double loop learning. © 2012 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of IPMA  Keywords : FSP approach; application similarity; fault inspection, detection and correction; quality assurance 1.   Introducing the program The client is a global company with investment banking and brokerage operations worldwide and is one of the world's largest financial institutions in the world. The client has embarked on an enterprise wide transformation program called Escape involving migration from Sybase estate to firm’s strategic databases MS SQL server and Oracle. The Vendor, one of the largest IT service and business solution  providers, partnered with client for end-to end activities involving assessment & program roadmap *  Corresponding author. Tel.: +912066086410; fax: +912066086399.  E-mail address : veerendrak.rai@tcs.com.  Available online at www.sciencedirect.com   © 2013 The Authors. Published by Elsevier Ltd.Selection and/or peer-review under responsibility of IPMA  409 Veerendra K. Rai et al. / Procedia - Social and Behavioral Sciences 74 ( 2013 ) 408 – 418  preparation, migration and de-commissioning of Sybase instances. Program Management unit of the vendor organization collaborated with other units of the vendor organization to execute this program. Authors of the paper are associates of the vendor organization. Program is being executed in Managed Services Mode (MSM) with a fixed price contract having commitments of progressive price reductions spanning over 18 months. This was a challenge and to overcome this and manage the program within the  budget, quality assurance activities were identified as one of the focus areas. Quality assurance is a not an add-on. It helps reduce time taken in identifying faults, preventing them from recurrence and reducing rework. Towards this, program management team developed and deployed FSP approach for defect management. 1.1.    Program background and business drivers Year 2008 spelt doom for many finance companies due to global economic meltdown. But, the client company took advantage of the situation to buy investment-banking and trading divisions of a financial company to significantly expand its earnings. This acquisition has brought many software applications into the fold of the client and increased the Sybase foot print. Early 2009 client realized that dependence on Sybase increased enterprise level risks as follows. •   Sybase does not have the revenue base to sustain R&D •   Sybase commercial strategy is to maximize revenue at the expense of longer term relationships. This prompted the client to rationalize the DB estate and identify MS SQL Server and Oracle strategic DB options for the enterprise in order to achieve the following benefits. •   Greater efficiency in engineering and operations, and •   Greater flexibility in cross-deploying development staff. 1.2.   The FSP Approach The traditional approach to defect management had heavy emphasis on fault detection. The process for this consisted of walkthroughs, inspection and various levels of testing. This lacked comprehensive approach to defect management with little focus on defect elimination leading to zero defects.  Also, the  similarities that exist in large migration programs are not considered in the defect prevention activities. Fig. 1. The FSP approach  410  Veerendra K. Rai et al. / Procedia - Social and Behavioral Sciences 74 ( 2013 ) 408 – 418 The best outcome of Defect Management is having zero defects by eliminating defects all together. This difficult task is targeted through a structured and comprehensive approach namely FSP Approach developed based on  Poka-yoke  Quality Assurance process of Toyota Production System (TPS). Invented in 1960s by Shingo (1997) a leading proponent of statistical process control (SPC), but disillusioned with SPC approach in which statistical sampling implies that some products go untested and defects will eventually reach the customer. Poka-yoke meaning  preventing inadvertent defects  is centered on the idea that defects are totally prevented to occur or they are early detected and corrected. Even though this technique was initially adopted in manufacturing domain, the underlying philosophy is readily applicable to software development as well. Schulmeyer (1990) and Tierney (1995) refer to Poka-yoke explicitly, but many other software quality experts have also championed detection and prevention methods in software. As Beizer (1990) observes in Software Testing Techniques, “We are humans and there will be bugs. To the extent that quality assurance fails at its primary purpose of bug prevention it must achieve a secondary goal of bug detection”. Expressing similar sentiment Steve Maguire (1993) observed in ‘Writing Solid Code’ that all of the techniques and guidelines presented in this book are the result of programmers asking themselves two questions: (1) How could I have automatically detected this bug? (2) How could I have prevented this bug? FSP approach lays emphasis on fixing the defects and preventing reoccurrence. Prevention is made  possible either by eliminating the defect completely through automated tools, utilities, checklists etc. or through early detection. This approach dovetails seamlessly into lessons learned process leading to continuous improvement of quality, which helps cost reduction. FSP helps take specific actions to detect the defects early in the life cycle (preferably in the same stage / step or subsequent stage / step). 1.3.    Double loop learning in FSP approach Fig. 2. Double Loop Learning in FSP approach FSP approach involves double loop learning with single loop learning at its core. In this program, migration work is performed and inspection of performed work is done to detect faults. Once faults are identified strategies are made to remove the faults. This is a standard cycle of inspection, detection and  411 Veerendra K. Rai et al. / Procedia - Social and Behavioral Sciences 74 ( 2013 ) 408 – 418 correction and involves single loop learning. FSP approach brings in another cycle of added learning wherein similarity dimensions are identified for applications being migrated, extent of similarities are assessed, causes are identified and cause defect ratio is determined. Strategies formulated in single loop learning are reformulated to exploit the similarities for defect detection, correction and prevention. Strategy reformulation involves developing tools, utilities, checklists and templates that exploit the similarities in the applications to be migrated and understanding the ‘content’ of the program better. Quality assurance does not come free and additional step involved in double loop learning in identifying exploiting similarities takes additional effort. However, this investment pays rich dividends later in the form of enhanced quality and effort and cost reduction. Table 1 describes the dimensions of similarities and Table 2 describes the extent / degree of similarities among applications to be migrated across the dimensions. Table 1. Dimensions of similarity Dimensions of similarities Includes Remarks Technology spread Core data base, application code, drivers connecting application to database Technological spread similarity is the most important dimension of similarity. The reason being that this program is about migration of applications from one data base to another. Business functionality Equity and Compliance are 2 types of the  business functionalities that applications implement. Business functionality is the added dimension which checks if the applications being migrated are in equity space or compliance space. Infrastructure Environment, Operating systems etc. Infrastructure similarity is least important in this context. Table 2. Degrees of similarities Percentage of Applications 40% 60% 50% Technological Similarity Business Function Similarity Infrastructure Similarity Extent of Similarity Overall Similarity 0.12 1 1 1 0.8 0.096 0.08 1 0 1 0.6 0.048 0.12 1 1 0 0.7 0.084 0.18 0 1 1 0.4 0.072 0.12 0 0 1 0.2 0.024 0.08 1 0 0 0.5 0.04 0.18 0 1 0 0.3 0.054 0.12 0 0 0 0 0.418 1 2.   Understanding the dynamics of FSP approach: System Dynamics Model System dynamics as discussed in Sterman (2001) and Mohapatra et al.  (1994) is an approach to understand non-linear behavior of complex systems over time. It applies to dynamic problems arising in any complex dynamic system characterized by interdependence, mutual interaction, information feedback, and circular causality. Non-linear behavior underlies all complex systems. Non-linearity cannot  be guessed intuitively and simulation is the only means to create and articulate this behavior. System  412  Veerendra K. Rai et al. / Procedia - Social and Behavioral Sciences 74 ( 2013 ) 408 – 418 dynamicists captured the informational representations of software development process modelling quite effectively. A thorough set of investigations of system dynamics modelling of software development  process can be found in the works of Abdel-Hamid and his colleagues (Abdel-Hamid & Madnick, 1986; Abdel-Hamid, 1989; Abdel-Hamid et al. , 1993, 1994). 2.1.   The Base Model FSP approach increases productivity by eliminating faults through prevention and early detection of errors. The base model captures the same through the following levels:- Work to be done; Work done correctly; Work with detected faults; and Work with undetected faults. Productivity increases through: Reduction of faults  and Early detection of faults. As the number of faults (defects) decreases through prevention the amount of rework decreases, while early detection decreases rework as well as the time taken to detect the faults and resolving them. Earlier the faults are detected less rework is done is established principle of project management. Figure 3 shows the core system dynamics model. Names of variables used in the model are given in Table 3. Fig. 3. The base system dynamics model
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks