1 PennDOT ATX Project Summer Semester EOSP Team Stalagmite: Dan Abramovich Jeff Ditillo Oksana Schubert Alexey Stolpovskikh Dehua Zhang
2 Agenda Introduction Planning Execution –Architecture-driven process –Project tracking –Risks mitigation Product –Product design –Demo –ATX business value Lessons Learned
3 Introduction Problem description –New solution for PennDOT ’ s on-line vehicle title and registration services Major stakeholders –PennDOT –ATX (our clients) –Participating companies Business drivers –Improve PennDOT ’ s business processes –Allow ATX to become PennDOT ’ s Business Partner
4 Context Diagram
5 Major Accomplishments Delivered software solution –Core product functionality –Designed for future extension –Designed to meet Gatekeeper certification –Client: “This is exactly what we wanted” Realized architecture in implementation –Highest priority quality attributes fulfilled Effectively planned and tracked project Created effective software development process
6 Development Process Individual development of work items –Units correspond to architectural components –Quality through design/code reviews Nightly builds and regression testing Architecture monitoring Periodic integration tests –Via Test UI and test cases Track progress against project plan closely
7 Planned Schedule
8 Actual Schedule
9 Time Estimation Accuracy Estimated task time –Actual: 158 % Adjusted estimates –Actual: 86 %
10 Scheduled vs. Unscheduled Effort Summary Scheduled task time –Tasks from project plan Unscheduled tasks –Role activities –Meetings –Defect correction –Miscellaneous
11 Time Expenditure Summary
12 Defect Tracking Summary Defects found using –regression testing –design/code reviews –Integration testing 53 Resolved 3 Unresolved –1 Enhancement –1 JBoss issue –1 Not Repro.
13 Anticipated Risks Untimely access to information and resources schedule slips Schedule is based on “guesstimates” may be unrealistic Only suboptimal choices available with free software
14 Risk Management Risk 1: Unreliable information access –Developed Gatekeeper simulators –Externalized critical data Risk 2: Estimated schedule –Develop transaction system kernel first –Project tracking worked to schedule Risk 3: Open source quality –Used proven products: JBoss, Jasper, Castor, Ant, JUnit
15 Implemented Architecture
16 Quality Attributes Achieved Security –J2EE JAAS (Java Authentication and Authorization Service ) Availability –Application and database clustering –Client-side dynamic proxy make fail-over transparency Modifiability –Decoupled business rules from the code using XML files Performance –Scalable load balancing
17 Security Features Username/Password authentication Role-based authorization Content-based authorization Security modules –Future extensions New security settings take effect instantly
18 Demo: Availability
19 ATX Business Value ATX transaction server –Provides key component of PennDOT business partner solution –Scalable solution –Easy to add additional transactions –Easy to add additional reports –Provides solution to update fees –Competitive advantage for Gatekeeper Certification Useful documentation –Developer’s Guide for future developers –System Administrator’s Guide
20 Lessons Learned Realistic project scoping is critical Selection as design Project planning/tracking –Identify and meet essential deliverables Use architecture driven approach Development process –Utilize automated tools –Need checks to ensure quality process Requirements/project management –Involve relevant stakeholders (PennDOT)
21 Q&A
22 Additional Slides
23 Scheduled Time Totals
24 Unscheduled Time Totals
25 Detailed Design: Transaction Management Transaction Manager –Transaction Record –Fee Management –Activity Log Transaction Descriptions Gatekeeper Interface