COMPANY: Union Bank, N.A.

CATEGORY: Superior Alternatives

INDUSTRY: Financial Services

GEOGRAPHY: Headquarters: San Francisco, CA

BUSINESS CHALLENGE: An aging and costly IT infrastructure was impeding the ability of Union Bank to scale to growth and respond agilely to changing market dynamics

MIGRATION PATH: UNIX™ on high-end RISC machines to Red Hat Enterprise Linux running on Intel Xeon based HP servers; Websphere to JBoss Enterprise Application Platform.

SOFTWARE: Red Hat Enterprise Linux™, Red Hat Network Satellite, JBoss Enterprise Application Platform™, JBoss Seam, JBoss Hibernate, Red Hat Consulting

HARDWARE: More than 150 Intel™ Xeon™ processor-based HP ProLiant servers

BENEFITS: Improve reliability and scalability, cut costs, and deliver new financial services and products to market faster

Download the PDF case study

BACKGROUND
Union Bank, N.A., headquartered in San Francisco is a full-service commercial bank providing an array of financial services to individuals, small businesses, middle-market companies, and major corporations. Union Bank is California’s fifth-largest bank by deposits. The bank has 335 banking offices in California, Oregon, and Washington and two international offices. Its holding company, UnionBanCal Corporation, is the 16th largest commercial bank holding company in the U.S. based on assets at March 31, 2009.

Union Bank was selected for its operating platform migration from AIX to Red Hat Enterprise Linux and Websphere to JBoss to support its mission critical applications at an improved price with greater performance and less up-keep. Union Bank used open source solutions to increase time to market, reliability and return on investment.

BUSINESS CHALLENGE
When Mok Choe joined Union Bank in early 2007 as chief technology officer, the Union Bank IT infrastructure faced a host of challenges similar to those of many other companies at the time, mainly increasing costs and resources associated with the maintenance and upkeep of legacy systems.

Over the years, Union Bank’s IT infrastructure had grown increasingly large, cumbersome, and complex. Not only was it costly to operate and maintain, but it couldn’t scale to accommodate the bank’s rapid expansion into new markets. System availability was also a continuing challenge. And as the financial services industry expanded into electronic banking products, Union Bank’s reliance on IT was increasing. The bank thus required an IT infrastructure that could speed new products to market with rock-solid reliability and availability, and which could also scale as needed.

The hardware environment embraced a “big box” approach with a few massive servers at strategic locations that offered little relief when significant impacts occurred. This environment required tremendous overhead with constant monitoring and management of server problems.

The IT department at Union Bank was also under pressure to reduce the total cost of ownership (TCO) of its overall IT operations. The solution needed to deliver a robust disaster recovery environment with minimal mean-time-to-restore (MTTR) and maximum mean-time-between-failures (MTBF) times. Finally, the solution needed to better leverage Union Bank’s most highly skilled IT workers. By enabling valued staff workers to reduce the day-to-day support required by overhead-intensive legacy systems, productivity would improve, and the bank’s IT department could move from a reaction to proactive support model.

“First and foremost, we needed to improve system availability,” said Choe. “Secondly, we needed to speed time to market of new financial services products. And at the end of the day, we needed to decrease the cost per transaction of delivering services.”

SOLUTION
Union Bank immediately focused on the task of establishing a new and innovative technology environment. The first decision: to create a new open source-based enterprisewide IT platform to obtain improved availability, agility, scalability and reduced TCO (total cost of ownership), while enabling the support of the bank’s growing IT needs and better alignment with the bank’s overall business plan.

“We did three specific things,” said Choe. “First, we migrated our entire Web-based infrastructure over to Red Hat Enterprise Linux so we could go from a scale-up to a scale-out architecture. Next, we ported our teller platform over to JBoss. And third, we wrote a brand new Web-based cash management application built on the entire Red Hat technology stack: Red Hat Enterprise Linux, JBoss, Hibernate, and SEAM.”

The strategy started at the operating platform level by replacing the aging UNIX based RISC servers with commodity x86 machines running Red Hat Enterprise Linux, and migrating to JBoss Enterprise Application Platform at the application server level. Union Bank initially utilized Red Hat Network to set up centralized, secure management of its Red Hat Enterprise Linux systems.

Union Bank took advantage of Red Hat Consulting to assist the IT group with the initial design of the first phases of deploying the new architecture and Web-based applications. The bank’s infrastructure and application development teams attended Red Hat Training to learn valuable tools and lessons on integration and migration issues.

The new strategy also encompassed building a new data center that leveraged virtualization technology on top of Red Hat Enterprise Linux to dramatically reduce the bank’s hardware footprint. “The bank is very serious about its green initiative, and Red Hat Enterprise Linux is a key part of that,” said Choe.

One of the most strategic projects was to replace the bank’s operating system environment on branch teller systems with JBoss Enterprise Application Platform running on Red Hat Enterprise Linux. Within just months, the Union Bank development staff was able to create a “silent” JBoss deployment package and distribute it remotely to over 330 production branch servers.

“The JBoss-based teller application has been running successfully at the 330 branch sites ever since,” said Choe, “The small footprint of JBoss has freed up much needed space on each branch server and has laid the ground work for future expansion. We plan to migrate other customer-facing web applications from Websphere to JBoss Enterprise Application Platform.”

BENEFITS
Union Bank’s innovative approach to its IT re-architecture has resulted in improvements to system availability, scalability and, resiliency, increased ROI, enhanced security, provisioning, configuration management, and improved time to market.

The most significant benefits have been improved system availability and resiliency. Upon migrating to Red Hat Enterprise Linux, there have been improvements of the bank’s hardware infrastructure, as seen by improved mean-time-to restore (MTTR), and mean-time-between-failures (MTBF).

The return on investment (ROI) was also substantial. For example, the large RISC machines were running at less than 50 percent capacity. To ensure redundancy, the bank needed to double its hardware investment to allow for fail over. “With Red Hat’s commodity model, we were able to spread the load over multiple machines and reduce our overall spend by approximately 80 percent,” said Choe. “And these savings don’t take into account the reduced maintenance costs of moving to the Red Hat platform, which is easier – and therefore cheaper – to maintain.”

Additionally, because application performance increased significantly under the new JBoss and Red Hat architecture, the bank was able to reduce the time-to-market of new products. The bank was also able to improve customer service by boosting the performance of its teller application. “The success of that project gives us confidence to tackle the rest of our browser-based Web applications with a JBoss solution,” said Choe.

The move from a vertical to a horizontal architecture and process enhancement have improved both system availability and resiliency, which allows the bank to absorb normal glitches without impacting customer transactions. “The reliability of our Web applications has improved to the point where I can go to our business partners and confidently say we have better than ‘four 9s’ availability,” said Choe.

The Red Hat/JBoss solution requires less maintenance and enables Union Bank IT to reduce their efforts on day-to-day support of legacy systems, allowing for better resource utilization. This also helped the IT group move from a reactive to a proactive model more expediently.

Additionally, the bank’s overall cost-per-transaction declined 25 to 40 percent, something that Union Bank’s business centers appreciate. “We have a charge-back system in which our departments pay for the IT resources they consume,” said Choe. “They’ve seen their charges go down month by month.”

“We benefited greatly from Red Hat consulting services as they provided valuable input and assistance in helping us migrate to Red Hat technology and dramatically improved our ability to achieve our goals,” said Choe, “With Red Hat Consulting, we felt there was an immediate knowledge transfer, and we were very satisfied with the level of involvement and quality of knowledge provided to our team.”

And ultimately many of the ongoing benefits that Choe expects to reap in coming years as a result of transforming the bank’s IT operations come from his expanded technology options. “We’ve achieved tremendous cost, reliability, and availability benefits, but in the end it all comes back to the fact that we now have choices when it comes to deploying hardware and software,” he said. “We’re no longer locked into using a particular product or vendor. Open source – and by extension, Red Hat – makes that possible.”

“The high costs and overhead associated with legacy proprietary-software and infrastructure led us to the decision to deploy Red Hat and JBoss open source solutions, and this allowed us to provide core infrastructure and development platforms at a significantly lower cost and at a faster rate,” said Choe, “Our use of Red Hat and JBoss solutions demonstrate creative business innovation through the use of horizontal architecture and the improvements allow Union Bank to continue to increase our customer experiences.”

COMPANY: GEICO (Government Employees Insurance Company)

CATEGORY: Superior Alternatives

INDUSTRY: Insurance

GEOGRAPHY: US

BUSINESS CHALLENGE: Existing proprietary middleware platform was complex to manage, not performing and scaling as expected and expensive to maintain. The architecture team decided to investigate alternatives that could be deployed that would better meet their needs.

MIGRATION PATH: Proprietary middleware platform to JBoss Enterprise Middleware

SOFTWARE: JBoss Enterprise Application Platform: 28 bands (1 band = 32 CPUs), JBoss Technical Account Manager (TAM), Red Hat Consulting, Amentra

HARDWARE: 50 Dell servers

BENEFITS: Reduced the total cost of ownership by more than 30%, throughput gain of 3X with utilization down to 1/3rd of the current platform, overall resource utilization went from above 50% to under 10% which allowed significant room for scalability without having to acquire additional hardware.

Download the PDF case study

COMPANY BACKGROUND
GEICO (Government Employees Insurance Company) is the third-largest private passenger auto insurer in the United States based on the latest 12 months written premium. GEICO provides auto insurance coverage for nearly 9 million policyholders and insures more than 14.4 million vehicles.

In addition to auto insurance, GEICO also offers customers insurance for their motorcycles and homes. Commercial auto insurance, boat, ATV, RV, personal umbrella protection and life insurance are also available.

GEICO is a wholly-owned subsidiary of the Berkshire Hathaway group of companies, is rated A++ for financial stability by A.M. Best Company and ranks at the top of several national customer satisfaction surveys. For more information about GEICO, go to www.geico.com.

BUSINESS AND/OR TECHNICAL CHALLENGE
In 2007, GEICO’s enterprise architecture team recognized they were facing several challenges with their existing proprietary middleware platform. The platform was complex to manage, not performing and scaling as expected and expensive to maintain. The architecture team decided to investigate alternatives that could be deployed that would better meet their needs.

The GIECO IT team identified the following challenges with their existing proprietary solution:

- Cost – GEICO’s license agreement was a “time bound licensing agreement” related to the number of proprietary application servers deployed during the time frame. Since GEICO experienced significant growth during this time frame, the cost to “true up” and pay for the additional licenses was significant.

- Performance – When GEICO upgraded their standard Java Development Kit (JDK) from version 1.4 to 1.5 on their existing proprietary platform, they did not see any improvements in machine (CPU/Memory) usage or application response time. After eight weeks of performance testing and tuning, they were finally able to configure the upgraded proprietary platform to match the earlier version’s performance. The upgrade was not only cumbersome but was also expensive since they had to engage external consultants to accomplish the upgrade.

- Memory leaks – The previous proprietary deployment also experienced unexplained memory leak(s). Developer load and memory testing returned misleading results unless the developer knew how to work around the leaks and complete certain types of tests.

- Documentation/Support – GEICO found it challenging to identify and understand the Java API in the current proprietary environment due to lack of documentation. They also had challenges in acquiring tools to identify memory issues, debug leaks, etc. For every instance of a high severity issue such as memory leak, external consultants needed to be engaged to identify and fix the problem.

- Staging – Due to these challenges, some of the GEICO development teams adopted JBoss technologies for their developer workstations and began building applications using JBoss. This dual use strategy became complex and redundant for IT Operations as they needed to make configuration changes on both the proprietary and JBoss platforms.

VENDOR SELECTION PROCESS
GEICO conducted extensive research and identified Sun’s GlassFish and Red Hat’s JBoss Enterprise Middleware as potential solutions that were suitable for GEICO’s application and infrastructure. JBoss Enterprise Middleware was selected based on its’ market share and extensive support from Red Hat. GEICO conducted a proof-of-concept, installing JBoss Enterprise Application Platform in a cluster of servers (POC environment). Performance and load tests were conducted using various tools for a selected business application on both platforms.

The JBoss results from these tests were astonishing. A few highlights include:
- User page transition time decreased as much as 19 seconds using JBoss

- During the proof-of-concept 1,749 additional business processes were created on the JBoss platform

- On the same hardware and environment, JBoss required 70% less CPU resources than the current platform

- Performance tuning with JBoss was accomplished in 40 man hours versus 1440 man hours for the existing proprietary platform

GEICO also conducted multiple reference checks with organizations that were of similar size and industry. The reference checks were extremely positive about Red Hat and JBoss Enterprise Middleware.

SOLUTION
The solution consisted of subscriptions for JBoss Enterprise Application Platform (EAP) and the initial deployment environment consisted of 540 processors. An additional 350 were added at a later date. A plan was put together to aggressively migrate 2 out of 3 mission critical applications in a time span of 3 months. GEICO also utilized a JBoss Technical Account Manager (TAM) who was dedicated to supporting GEICO’s specific needs during their switch to JBoss.

BENEFITS
By implementing JBoss Enterprise Application Platform, GEICO was able to reduce the total cost of ownership by more than 30%.

When compared to the previous proprietary platform, GEICO also experienced throughput gains of 3x, and a 2/3 reduction in utilization. The overall resource utilization went from above 50% to under 10% which allowed significant room for scalability without having to acquire additional hardware.

RED HAT SUPPORT, TRAINING, AND CONSULTING SERVICES LEVERAGED
One of the challenges for GEICO was the time bound migration process. GEICO’s middleware team was trained on JBoss for a week. With the support of Red Hat and Amentra (a Red Hat company), they successfully migrated the initial 2 applications and were able to migrate the 3rd application as well. This was a clear demonstration of expertise in Red Hat Consulting services and the ability of GEICO’s middleware team to adapt rapidly to the new JBoss environment.

ADVICE FOR OTHER COMPANIES FACING A SIMILAR BUSINESS CHALLENGE
“Open-source does not translate to unsupported. Don’t be afraid of change. GEICO had initial concerns about support, stability and deploying open-source software for its mission critical applications, but the market maturity and the premium level of enterprise support offered by Red Hat made it very easy to make the change to an open source environment. If your organization has been slow to consider adopting open-source solutions, they may lose a competitive advantage that can be gained based on lower cost of ownership and utilization of efficient/best of breed open source products.

COMPANIES: Massachusetts Convention Center Authority (MCCA) and Optaros

CATEGORY: Extensive Ecosystem

INDUSTRY: Convention center management

GEOGRAPHY: Boston, MA

BUSINESS CHALLENGE: Manual processes and siloed systems resulted in inefficient workflows that caused customer service to suffer.

MIGRATION PATH: From a Microsoft™ Windows – based client-server application to service oriented architecture (SOA) J2EE application based upon the JBoss Enterprise Application Platform

SOFTWARE: JBoss Enterprise Application Platform, JBoss Enterprise SOA, JBoss Messaging, JBoss ESB, JBoss jBPM, J2EE™, Google Web Toolkit™, Apache

HARDWARE: Intel Xeon™ – based x86 servers

BENEFITS: More efficient and timely access to data and an automated streamlined workflow that improved worker productivity and customer service levels and increased revenues. Approximately 90 percent of the MCCA’s day-to-day operations are run using open source technologies.

Download the PDF case study

BACKGROUND
The Massachusetts Convention Center Authority (MCCA) owns and oversees the operation of four major facilities, including the Boston Convention & Exhibition Center (BCEC). The MCCA’s mission is to generate local economic activity by attracting conventions, tradeshows, and other events to its world-class facilities. The MCCA has generated $2.3 billion in economic impact over the past five years in the greater Boston area and is the eighth busiest convention center in North America.

Optaros is an open source consulting firm that specializes in the development of custom applications for clients through the Assembled Web. Optaros and MCCA staff worked together to rebuild the MCCA’s event management system. The product, known as ShowBiz, helps streamline the detailed process of setting up large-scale events, and is now completely run on an open source stack that was designed and assembled by Optaros. Approximately 90 percent of the MCCA’s day-to-day operations now run on open source solutions from Red Hat and JBoss.

BUSINESS CHALLENGE
The MCCA’s business model is based on the premise that it has an “inventory” of convention center space to sell – space that is used to host trade shows, association meetings, and other events. And despite being owned by the state government, the MCCA prides itself on operating like a for-profit business for the economic benefit of the City of Boston and Commonwealth of Massachusetts. In particular, the MCCA team strongly believes in providing an exceptional client experience, which means being able to rapidly configure its facilities and tailor its services to deliver whatever a particular customer wants.

To do this, the MCAA required an IT structure that was fluid, flexible, and scalable. But the application the MCCA was using to manage the sale of space was antiquated and difficult to use. The biggest challenge was that personnel did not have ready access to data. Without this data, users couldn’t make the kind of smart, real-time decisions needed to optimize service delivery and revenues during each stage of an event’s lifecycle. Important business decisions were being affected by the antiquated software, often relying on intuition and guesswork rather than facts. Everyone in the organization was affected by the system’s faults – from the executive suite down to contract electricians.

For example, because the financial database had been separated from the event management system in 2006, all accounts receivable and event creation information had to be manually entered in both applications, creating duplicate work for all involved as well as introducing errors into the databases. Additionally, managing service delivery to the large number of exhibitors was primarily a paper-based manual process that didn’t support online ordering and payment. And then there was scalability. The MCCA realized that its systems were a serious impediment to its ability to grow as planned.

Because of the MCCA’s unique, multi-facility business structure and complex business processes, a commercial off-the-shelf application wouldn’t do. MCCA senior executives knew they needed a new, custom-built solution that was designed to support the specific needs of their business.

Moreover, such a solution needed to provide MCCA employees with more efficient, timely access to data. It had to automate workflows. It had to minimize manual processes and eliminate redundant data entry and utilize technologies that would provide flexibility and scalability for the future needs of the business. Finally, it had to deliver an easy-to-use and elegant user experience that the MCCA could eventually extend access to the application to clients and customers.

SOLUTION
After performing a thorough analysis of its needs, the MCAA brought in Optaros, a Boston-based professional services firm that designs, assembles, supports, and monitors custom Web applications using open source software. “Steve Snyder, the CIO and CTO of the MCCA liked the freedom and the choice that open source offered,” said Errol Apostolopoulos, a management consultant at Optaros, who managed the project. “The fact that we could build him an application based on industry standards was very attractive to him. There were the lower acquisition costs, of course. But then there was also the fact that all the open source technologies, tools, and platforms integrated together so well.”

The MCCA was looking for a scalable solution that could grow with the company — exponentially in regard to data structures. They determined that a SaaS, software as a service, type architectural model, allowing for plug-and-play, iterative updates would offer technology that could evolve and grow along with the organization.

From the beginning, it was clear that Red Hat’s JBoss Enterprise Application Platform was going to play a large role in the solution. “Part of our process is to go out and comb the open source community to find the best technologies we can leverage to build our solutions,” said Apostolopoulos. “JBoss was the absolute best choice for the MCCA.”

The MCCA solution contained three pairs of JBoss instances. The first pair was used for the online customer-facing site. By hosting JBoss Enterprise Application Platform in a clustered environment, the MCCA allows exhibitors to purchase services and materials online. Previously, they had to fax in their orders, which then had to be entered into the old event management system manually.

The second JBoss pair also involved using JBoss in a clustered environment, and was used for the MCCA’s internal event management site. This new event management application allowed MCCA personnel to manage all aspects of the event lifecycle – from sales, to event and space setup and configuration, to exhibitor services, to all financial aspects of the event. This application uses JBoss jBPM as the workflow engine for the initiation, review, and approval of space booking throughout the sales cycle, from pre-sales through confirmation upon receipt of the signed contract.

Optaros selected Google Web Toolkit (GWT) as the front-end of the system. Business services were developed using Hibernate frameworks to handle queries and transactions.

All JBoss applications and ESB servers were configured to run on Intel Xeon-based hardware under Windows Server 2003. The applications run on a cluster of SQL Server database servers configured for replication and failover. Today, the MCCA employs 10 production servers; four servers for quality assurance (QA) and testing; and two for developing enhancements to the system to run the application.

BENEFITS
MCCA personnel now have ready access to real-time data, as opposed to running reports and requesting information that was often hidden within the old system. Streamlined processes enabled by the new architecture have allowed staff to redeploy time previously spent on unnecessary manual and paper processes to focus on customer service. Overall, employees are much more efficient, and the corporate culture is much more customer-centric than under the previous system.

This tiered architecture is not only secure, but is also scalable, reliable, and available. For instance, the deployment manager can scale the three clusters independently based on their respective usage in terms of number of concurrent users, transactions volume, and more. From a security perspective, the ESB servers act as reverse proxies to a back-office financial management system and PayPal’s credit card processing network.

The fact that Optaros designed the applications using service oriented architecture (SOA)-based plug-in/plug-out framework means that the MCCA’s own IT team can integrate other external services into it as needed going forward. This gives the MCCA the flexibility and scalability to meet its growth objectives while keeping the main application stable.

From a financial perspective, the new applications have been a tremendous success. Employees are no longer wasting time manually inputting duplicate content into multiple systems, but can focus on higher-level tasks. As a direct result of this, the MCAA has been able to collect more than $500,000 in outstanding accounts receivables over the past six months.

And the new applications have allowed the MCCA to deliver a premiere customer experience. Under the old systems, work orders were comprised of 50- to 100-page documents that included details such as the number and location of chairs, the timing of food service, electrical needs, and everything else that impacts the success of an event. Today, all data related to an event is searchable, and MCCA customers are now able to order and update space, tables, internet access, electrical outlets and other services online rather than using the outdated paper faxing process. MCCA personnel are then electronically notified when there are any changes to work orders that affect their roles in an event, and the event system can be trusted to contain the most recent information.

These conveniences are only the first step. The MCCA wants to extend accessibility and transparency to the applications even further. For example, the MCCA hopes to eventually give taxi drivers access to the system so they can see in real time the transportation needs of people attending an event.

Finally, credit card processing is now completed automatically and in real time, instead of manual batch processing at the end of each business day.

“The competence of Red Hat’s consultants and support personnel clearly contributed to the application development team’s overall success,” said Apostolopoulos. “Their support enabled the project team to deploy systems more effectively with the assurance that additional assistance was only a phone call away. The Red Hat team went above and beyond our expectations.”

Apostolopoulos said that he recommends JBoss to its customers whether they are building new applications from scratch or migrating existing applications from proprietary hardware and software to an open source platform.

“We’re finding more enterprises choosing open source – not just because of its low cost and ability to scale, but also because of the flexibility it gives them to choose components that plug and play into their systems as their needs change,” said Apostolopoulos. “And JBoss is clearly the industry middleware standard for these increasingly strategic open source projects.”

“Our customers will soon have access to the same data the staff does, so people can order more services directly through the system,” said Steve Snyder, chief information officer for the MCCA. “The shopping cart and credit card processing for basic client needs are only the first step in offering more accessibility and transparency for customers to directly access data. The MCCA has hopes of allowing more constituents to access pieces of the system. The technology and system that was built, the cooperation between the JBoss, Optaros, and MCCA teams, the full buy-in from MCCA executives to end-users, and everyone being involved in the whole process truly made this deployment a resounding success.”

COMPANY: American Family Mutual Insurance Company

INNOVATION AWARD CATEGORY: Management Excellence

INDUSTRY: Property and Casualty Insurance

GEOGRAPHY: North America

BUSINESS CHALLENGE: Needed a cost effective centralized management solution for its 1,200 instances of JBoss Enterprise Application Platform that would scale with its growing computing infrastructure

MIGRATION PATH: ad hoc monitoring solutions to JBoss Operations Network

SOFTWARE:JBoss Operations Network (JBoss ON), JBoss Enterprise Application Platform on 150 plus systems, Red Hat Consulting, Oracle DB, HP OpenView

HARDWARE: Intel based Dell x86 commodity servers

BENEFITS: Improved availability and reliability of applications, a monitoring solution that can manage a large number of application server instances, simplified management, enhanced management and monitoring, and reduced costs

Download the PDF case study

COMPANY BACKGROUND
American Family Mutual Insurance Company, the nation’s third-largest mutual property and casualty insurer and 14th-largest property and casualty insurance company group, offers multiple insurance lines. These offerings include automotive, home, life, health, and business insurance.

American Family began with three employees in 1927 and has since grown to become a Fortune 500 company that generated $6.7 billion in revenue in 2008. American Family has 4,000 agents who serve 19 states.

BUSINESS/TECHNICAL CHALLENGE
An issue of scalability was American Family’s greatest technical challenge, as the Java application server environment was growing consistently and there was a need to identify a cost effective, stable and reliable management solution that would complement this growth. American Family’s prior management system could not scale to the size needed at an appropriate cost.

American Family deployed JBoss Operations Network (JBoss ON) across 1,200 JBoss application server instances and is using it to monitor and manage those systems.

DESIRED SOLUTION
American Family required a systems management solution that would provide: Real-time monitoring, alerting, historical trending, and the ability to control running systems in its JBoss Enterprise Application Platform environment. Performance of the product was a huge criteria-performance not only to scale, but to be responsive so that the operations team could use it successfully on a daily basis.

JBOSS PRODUCTS USED IN FINAL SOLUTION
About 3 years ago, American Family began migrating its IBM WebSphere environment to JBoss Enterprise Application Platform to run many of the company’s business critical applications, including billing and claims, customer information management, an agent-facing sales suite, and web services. As the JBoss environment grew, American Family quickly identified the need for the tools to help monitor and manage the servers. The company evaluated a number of tools from several leading vendors and selected JBoss ON based on cost, scalability, and functionality.

The American Family operations team was asked to monitor, manage, and control a very large computing infrastructure with several different tools. JBoss ON will allow the Computer Operations team to manage the entire JBoss infrastructure with one console.

American Family will also continue to look for additional opportunities to use JBoss ON with their application developers. It is hoped that by using JBoss ON, American Family will be better able to detect and fix problems earlier in the software delivery life-cycle.

BUSINESS IMPACT
American Family expects to see improved reliability in the application server space due to the JBoss ON feature of historical trend analysis of key system metrics and faster time to react when there are problems due to the alerts based on those same metrics.

American Family is hoping to extend this capability into application support areas to detect problems early in the software development lifecycle. Through alerting, monitoring, and the opportunity to proactively address situations before they cause an outage, JBoss ON will improve the reliability and availability of Java application server applications and keep internal customers satisfied.

Ad hoc management scripts and tooling will be replaced with JBoss ON. The result should reduce time and effort needed to manage, monitor and control systems.

VALUE-ADD TO BUSINESS FROM JBOSS
American Family has been working in a collaborative manner with the JBoss ON engineering, support and product team for close to two years. JBoss provided four people on-site and over the course of two days, they worked out a number of technical challenges that could not have been tested or seen in the lab. This collaboration has resulted in improved scalability and features.

END-CUSTOMER VALUE
By working together, American Family and Red Hat have given back something of value to the open source community.

RED HAT CONSULTING / SUPPORT
American Family worked closely with JBoss ON resources to ensure that the product met functional and non-functional requirements such as scalability and performance.

ADVICE FOR COMPANIES FACING SIMILAR CHALLENGES
American Family suggests detailing the desired solution’s requirements and identify a provider who not only can offer the specific product, but also the support and willingness to collaborate and devote resources to making the relationship successful.

INNOVATION
The American Family and JBoss collaborative work on JBoss ON blazed a trail for the future features and scalability of the monitoring and management product and will provide benefits to other customers and developers to capitalize on.”

Australian Insurance Company Saves $500,000 by Migrating from Windows and WebSphere to Red Hat Platform and JBoss Middleware Solutions, Gains Increased Flexibility, Scalability and Freedom

COMPANY:Allianz Australia Insurance Limited (Allianz)

CATEGORY: Carved Out Costs

INDUSTRY: Insurance

GEOGRAPHY: Australia

BUSINESS CHALLENGE: Rebuild Allianz’s IT infrastructure based on a flexible and scalable platform that could leverage new virtualisation technology to generate hardware and support savings, and reduce its underlying software and operations costs for several strategic business application projects

MIGRATION PATH: Windows and Intel-based infrastructure to Red Hat Enterprise Linux on IBM System z10 mainframe and from Websphere to JBoss Enterprise Middleware on Intel processor based HP ProLiant servers.

SOFTWARE: Red Hat Enterprise Linux, Red Hat Network Satellite, JBoss Enterprise Middleware, JBoss Operations Network, Red Hat Consulting

HARDWARE: IBM System z10 mainframe, HP ProLiant servers

BENEFITS: Significant reduction in middleware software and support costs, reallocation of IT budget from software licensing to staff and resources, resolution of data centre power limitations with new capacity for growth, superior workload management and operational efficiency, reduced carbon footprint, increased flexibility, scalability, and freedom from vendor lock-in

Download the PDF case study

BACKGROUND
Operating across Australia and New Zealand with approximately 3,300 staff, Allianz offers a wide range of insurance and risk management products and services, including personal insurance, industrial and commercial insurance, corporate insurance, public and products liability and workers’ compensation insurance.

Allianz is Australia’s fourth largest general insurer with over two million policyholders. It also provides some form of insurance cover for more than half of Australia’s top 50 BRW-listed companies (2005).

The organisation has been a member of the Australian Government’s Greenhouse Challenge Plus program and the wider Global Allianz Group has a target to reduce its greenhouse gas emissions by 20 per cent by 2012.

BUSINESS CHALLENGE

In 2007, Allianz Australia’s IT department reached a crossroads as it faced a data centre that was at capacity and network equipment that was at end of life. As a long-standing Windows shop, Allianz saw the situation as an opportunity to build a completely new infrastructure from the network right down to the back-up devices.

“After running WebSphere applications on an Intel platform using Windows for some time, we realised we couldn’t go any further with the current set-up,” said Peter Rowe Head of Infrastructure and Operations, Allianz Australia Limited.

In addition, Allianz had also reached the limitations of its Windows based operating platform.

“We needed a platform that could give us the flexibility and scalability to enable us to grow and expand for the future,” said Rowe.

In line with Allianz’s goal to reduce emissions by 20 percent by 2012, the IT department was also interested in examining how it could take advantage of new virtualisation technology to save on power usage.

SOLUTION
After comprehensive analysis, Allianz concluded that the most viable option for the business going forward would be to:

- combine the performance and reliability of the IBM System z10 mainframe with the flexibility and efficiency of Red Hat Enterprise Linux; and

- deploy JBoss Enterprise Middleware on the Intel platform – implementing a common system across diverse hardware

In September 2008, Allianz received project approval to commence both infrastructure refreshes concurrently. “IBM System z10 mainframe running Red Hat Enterprise Linux was undoubtedly the best fit-for-purpose solution for us,” said Rowe.

“The mainframe offered us the best option for consolidation – Our business took the view that the mainframe was essentially another commodotised piece of hardware and for us, the value really lies in the interface between the server and the applications. Our target was to employ a Red Hat Enterprise Linux based platform that could combine agility with low support costs.

“When assessing operating systems for the IBM System z10, Allianz found that open source based Linux would deliver the best outcome and as a result, selected Red Hat Enterprise Linux, to host all Internet-facing applications, including home and motor insurance quoting, broker and agency pages, premium funding pages, and other broker tools for policy servicing.

“A major part of the decision to use Red Hat Enterprise Linux was its tight integration with the IBM platform and its impressive support structure,” said Rowe, “Red Hat’s relationship with IBM meant that if we were going to have any issues, Red Hat could essentially replicate the scenario on their own IBM z10.” For ongoing, centralised systems management, Allianz also implemented Red Hat Network Satellite to complement the management of Red Hat Enterprise Linux and minimise daily administrative demands.

As a centralised tool, Red Hat Network Satellite can help boost productivity by creating a single template for managing multiple servers more efficiently. It creates a system for grouping machines together and mapping out how to manage them. Having this template also means that one administrator can run the same updates on a greater number of systems simultaneously, and can build a new and completely configured machine within a couple of minutes.

In addition, it has the capability to manage virtualised instances of Red Hat Enterprise Linux through the same interface regardless of underlying hardware platform. Red Hat Network Satellite manages Red Hat on the mainframe in the same manner and interface as Red Hat on a distributed machine.

As an alternative to WebSphere, which Allianz had been using as part of its existing framework, the organisation implemented JBoss Enterprise Application Platform for content management, document management, business process automation, and client portals, along with JBoss Operations Network Management to reduce systems management and resource costs.

“We had three new workloads – BPM, content management, and print services – that had begun on the old system, and which had to be redeployed in order for us to achieve the projected capacity needed for these workloads over the next two to three years,” said Rowe.

“JBoss has enabled us to cost-effectively leverage the recent advancement in Intel chipsets in HP hardware, without the costs traditionally associated with multi-core software licensing models,” said Rowe.

Allianz enlisted Red Hat and JBoss Consulting services to assist in building automated provisioning of new Red Hat Enterprise Linux and JBoss Enterprise Middleware deployments through Red Hat Network Satellite and JBoss Operations Network in a matter of minutes rather than days.

BENEFITS
Commencing production in late April 2009, Allianz has already observed a number of solution benefits during the testing phase.

“The combination of Red Hat Enterprise Linux, Red Hat Network Satellite, JBoss Enterprise Application Platform and JBoss Operations Network was, and will continue to be, valuable for Allianz, as the solutions enabled the organisation to streamline operations and free up resources for allocation to other high-value tasks,” said Rowe.

Allianz will save more than $500,000 in middleware licensing costs, as it deploys JBoss Enterprise Middleware on the Intel platform.

Plus, by using Red Hat and JBoss Consulting and Training, the Allianz team has been able to get up-to-speed in a very short period of time, once again freeing up funds to invest in areas such as staff development, rather than expensive proprietary infrastructure software.

“We were looking for an operating system that we could simply drop down into the network, that was independent of the hardware, and provided a higher level of service with centralised management for patching,” said Rowe.

“So far with Red Hat Enterprise Linux, Red Hat Network Satellite, and JBoss Enterprise Middleware, we’ve begun to significantly reduce day-to-day security administration and now have more strings in our bow in terms of virtualisation and what we can do to minimise both the investment in hardware and ongoing production costs – particularly those associated with escalating licensing and support costs,” said Rowe.

Allianz believes this is just the beginning, with expectations that the move from a Windows-based environment to a virtualised Red Hat Enterprise Linux environment, together with the JBoss Enterprise Middleware deployment will save the organisation over one million dollars a year in hardware and support costs.

In addition to resolving data centre capacity issues, the new system is expected to accommodate significant future growth whilst enabling superior workload management and operational efficiencies, and helping to reduce the organisation’s carbon footprint.

Allianz also believes it will continue to benefit from the fact that Red Hat has more applications certified to run on Red Hat Enterprise Linux than any other open source operating system platform.

Once the production system is up and running, Allianz will shift its focus toward the continued reduction of proprietary software in favour of more widespread use of Red Hat Enterprise Linux and JBoss Enterprise Middleware.

COMPANIES: Harvard Business Publishing (HBP) and Rivet Logic

CATEGORY: Optimized Systems

INDUSTRY: Publishing

GEOGRAPHY: Cambridge, Massachusetts

BUSINESS CHALLENGE: The HBP’s ability to get new products to market and the quality of the customer experience at its e-commerce site were hindered by a proprietary operating system, a difficult-to-use legacy content management system (CMS), and inflexible customer-facing Web applications, which were negatively impacting the HBP’s revenues and limiting growth

MIGRATION PATH: From a proprietary operating system running a proprietary legacy CMS application to Red Hat Enterprise Linux and JBoss Enterprise Application Platform running the Alfresco Content Management System

SOFTWARE: Red Hat Enterprise Linux, JBoss Enterprise Application Platform and Frameworks including JBoss Seam, JBoss Hibernate, jBPM, Oracle Database, and Alfresco’s open source Content Management System

HARDWARE: Intel™ Xeon™ processor-based Dell™ 2950 multicore servers

BENEFITS: Increased employee productivity, lowered IT operational costs, and increased Web site traffic and e-commerce transactions

Download the PDF case study

BACKGROUND
Harvard Business Publishing (HBP) is a not-for-profit, wholly-owned subsidiary of Harvard University which publishes a range of content – both print and online – bridging the knowledge gap between academic and the corporate world. It serves three primary markets: academic, enterprise, and individual managers. With more than 250 employees, the HBP’s mission is to explore and improve management practices around the world. HBP’s major Web properties include the online version of Harvard Business Review (hbr.org), Harvard Business Digital (harvardbusiness.org), and Higher Education (www.hbsp.harvard.edu). Rivet Logic and HBP were selected for the Optimized Systems Innovation Award for the use open source solutions from Red Hat, JBoss, and Alfresco that have enabled increased stability and the ability to develop products faster, bundle existing products more efficiently, and generate new revenue opportunities by increasing site traffic and offering richer, fresher, and more varied content.

Rivet Logic provides professional open source services and solutions that help organizations engage with customers, improve collaboration, and streamline operations. The company offers a full suite of JBoss professional services – including deployment, customization, and integration – enabling clients to fully leverage the power of the world’s leading open source enterprise middleware stack. With complementary expertise in the Alfresco content management platform, Rivet Logic offers integrated, content-rich, and Web-oriented architecture (WOA)-enabled solutions that power a new generation of interactive Web properties, enterprise intranet applications, and collaborative Web 2.0 communities.

BUSINESS CHALLENGE
To stay innovative and develop new products faster, HBP’s business users require the ability to easily access and use content from a variety of systems across the range of HBP business units. But the existing aging content management system was limiting access to only a few trained power users, which routinely resulted in productivity bottlenecks across all units.

To further challenge the workflow and production of HBP products, critical content resided on various shared drives across the enterprise or was locked up in the proprietary system, making it increasingly difficult for HBP to repurpose existing content into the kind of new digital media products that the fast-moving business information marketplace was seeking.

“Strategically, HBP knew it needed to transform itself from a print organization – which what it was for the past 10 to 20 years – to a digital media organization,” said Mike Vertal, CEO of Rivet Logic Corporation, a professional open source services and solutions firm hired by the HBP to reengineer the core IT platform and mission-critical applications.

The growing array of aging and disparate legacy middleware and operating systems used to run HBP’s Web sites was also proving increasingly unstable. The system routinely caused integration hurdles, IT bottlenecks, and escalating operational costs due to personnel overhead and software licensing fees. The lack of easy-to-use Web publishing tools hindered the editorial staff’s ability to deliver fresh and innovative content and, consequently, limited HBP’s ability to drive site traffic and therefore the ad revenue and e-commerce transactions that contributed directly to the firm’s bottom line.

In addition to the financial overhead due to high software licensing and maintenance costs, a large percentage of IT operational costs and human resources were spent just keeping the old systems running, leaving little time and resources for developing innovative new products. The proprietary legacy systems were difficult to customize and integrate, and could not scale to keep pace with HBP’s expanding business.

SOLUTION
HBP recognized it needed to replace its proprietary content management system with a robust, yet easy-to-use enterprise-grade content management system that would facilitate access to its high-value content to its business users and integrate seamlessly with existing systems such as enterprise content repositories, search and merchandising tools, e-commerce systems, ad networks, Web analytics, and community-building applications such as blogs.

HBP required a solution that provided increased flexibility around page design and messaging, easy access to digital products, a uniform user experience, easy-to-use e-commerce experience, and improved visitor experience for user registration and session management. HBP also sought a higher level of performance, scalability, and rock-solid stability.

One absolutely non-negotiable requirement: the new solution needed to be built with open source software and an open architecture with an enterprise Java foundation at the core. It also needed to support rapid, lightweight development at the upper layers of the application stack – most notably at the user interface layer and presentation tier. This requirement would focus on HBP’s business goals and on leveraging HBP’s very high-value content and core capabilities to enable future innovation.

This is where Rivet Logic came in. Rivet Logic provides professional open source services and solutions and offers a full suite of JBoss professional services including deployment, customization, and integration – enabling clients to fully leverage the power of the world’s leading open source enterprise middleware stack.

Rivet Logic implemented an end-to-end open source solution that delivered on all of HBP’s requirements. HBP’s production ecosystem was built on Red Hat Enterprise Linux on Intel Xeon processor-based Dell 2950 servers with dual and quad core CPUs, running JBoss Enterprise Application Platform, Apache, Oracle Database, and the Alfresco Content Management System.

JBoss Enterprise Application Platform was used as a core component for the dynamic content delivery system and e-commerce experience. JBoss Enterprise Application Platform provided the basis for a WOA that enabled straightforward integration with numerous enterprise back-end systems and third-party Web services, including enterprise resource management (ERP), ad servers, XML repositories, taxonomy management, third-party search, Web analytics, and a user ID management system.

In addition, JBoss Seam served as the rich user interface (UI) framework for an intranet application for enterprise content management, and the public-facing Web applications for the online versions of Harvard Business Review at hbr.org, HBP’s e-commerce site at harvardbusiness.org, and HBP’s Higher Education site at www.hbsp.harvard.edu. In all cases, the JBoss Seam applications were integrated with Alfresco for back-end content management. The intranet application utilized Alfresco’s document management (DM) repository, whereas the Web site applications utilized Alfresco’s Web content management (WCM) repository.

JBoss Hibernate provided the persistence layer for all application logic and user-generated content, and jBPM governed workflow for editorial content and publishing processes. The JBoss Enterprise Application Platform provided the foundation necessary for HBP’s mission-critical applications that required high performance and scalability. Rivet Logic used a WOA approach for the need for single-sign-on support, while also enabling integration with a variety of systems, including a blogging platform (blogs.harvardbusiness.org), e-commerce, an XML repository (for HBR article content), and community platforms. Integration with a third-party search engine offered powerful faceted search and navigation functionalities. This content delivery approach also met standards-compliant XHTML/CSS requirements, maintained SEO-friendly URLs, and allowed for straightforward integration of Web analytics. Integration between JBoss and Alfresco was streamlined by using free and open source software from Rivet Logic, including the Remote Alfresco API rivet for Alfresco DM integration and the Crafter rivet for Alfresco.

The JBoss Seam intranet application allows enterprise users to:

- Navigate, search, find and retrieve relevant content quickly through a combination of full-text search, metadata search, and content relationship browsing

- Create and enter new content and associate metadata and relationships

- Manage digital rights of product-related media

- Restrict access to certain types of content through role-based user authorization

“The new JBoss and Alfresco based intranet provides an easy way for end-users to search and find content, as the search results deliver detailed content, such as individual chapters, images, author bios and the public-facing HBP site provides visitors a rich experience for navigating and consuming HBP’s digital content,” said Vertal, “The JBoss and Alfresco based Web content delivery system provides the dynamic and feature-rich functionalities HBP needed in a simplified manner by seamlessly connecting the presentation, application and content repository layers.”

BENEFITS
By using Red Hat, JBoss, and Alfresco open source solutions, HBP has gained platform agility that enables brand management, broader community functionality, and increased site traffic. The JBoss and Alfresco integrated solution has enabled HBP to gain the stability and ease of use it required to empower business end users and integrate with a host of critical applications and systems. With the new system in place, HBP can now develop products faster, bundle existing products more efficiently, and generate new revenue opportunities by increasing site traffic and offering richer, fresher, and more varied content.

From a developer perspective, HBP’s IT department can now focus on value-added development of new application and site features given the open source architecture and the modern WOA infrastructure. Dramatically less time and resources are now spent on maintaining rigid, legacy systems that carried expensive maintenance and software licensing costs.

The new implementation has enabled HBP to better leverage the value of its branded content, including articles, books and book chapters, blogs, podcasts, and videos – easily, quickly and securely.

“Uniting all content across the enterprise led to the rapid development of new digital media products and richer content on HBP’s revenue-generating Web properties,” said Vertal, “With Red Hat, JBoss and Alfresco, HBP has enhanced the visitor experience with improved navigation, along with much faster Web site performance. By offering fresher and more dynamic content and increasing site traffic, HBP has started to expand its revenue opportunities.”

“This project was as mission-critical as they come,” said Vertal. “It encompassed the back-end repository, the front-end application that internal users deployed to create new content and products, and a customer-facing Web application that delivered those products to customers through a variety of channels. Red Hat Enterprise Linux coupled with JBoss Enterprise Application Platform drove a total transformation of the way that HPB approached product development and delivery.”

Business agility has also increased by orders of magnitude, said Vertal. “Because we were on the new platform, in a matter of months we were able to replace the entire e-commerce front end with a much better user experience and more manageable applications.”

The stability of the system has also proven itself. And, looking forward, HBP has plans to begin incorporating social media into the site using collaborative tools and community applications that will enable its employees to become better engaged with customers. “This will allow HBP to build and maintain better customer loyalty across its entire customer base,” said Vertal.

“We utilized leading edge, open source platforms from Red Hat, Alfresco, and Rivet Logic to implement a large-scale, high-value, business-critical solution that spans internal enterprise collaboration applications, public-facing Web properties and communities, and business-critical e-commerce applications,” said Vertal, “We believe this project demonstrates the powerful benefits that commercial open source software from Red Hat, JBoss and the open source ecosystem is ready to deliver to major enterprises for the converged world of content, community, collaboration, and commerce.”
CUSTOMER ADVICE

“Any enterprise or government agency that is looking to increase employee productivity and/or improve relationships with customers should seek to leverage next-generation solutions that expand their use of content, community, collaboration, and community. And just as the consumer Web 2.0 was built on open source software, these next generation Enterprise 2.0 solutions are being built on enterprise-grade, commercial open source software from Red Hat, JBoss, Alfresco, and others. All organizations should seek to leverage commercial open source software as much as possible for any and all future enterprise software initiatives,” said Vertal.

“Businesses should remember that software is a means to an end, not an end in itself. Companies should first focus on business requirements and desired results, and leverage the best-of-breed software technologies that will help them get there. And whether the business needs better internal-facing, content-enabled enterprise applications, improved external-facing Web properties, or e-commerce platforms, JBoss software has proven it can help deliver tremendous bottom- line results,” said Vertal.

Category: Certified Service Provider

Winner:Amentra

Submitted by: Amentra Team

Industry: Amentra = Technology / La Petite Academy = Education

Geography: Virginia

Overview
Selected for helping enterprises deploy mission-critical business systems on JEMS through a formal, experience-proven mentoring and software development program.

Download JBoss Innovation Award Submission
Read Amentra Case Study

1. Please describe your company. (Number of employees, private/public, industry, etc.)
Amentra, Inc. offers a distinctively different approach to business and IT consulting. By helping clients deploy mission critical business systems through a formal, experience-proven mentoring and software development program, Amentra has earned industry accolades for combining two areas that have historically been separate service offerings into a single solution: deliverable-based project solutions integrated with IT Mentoring. Amentra has great expertise in retail, insurance, pharmaceutical, telecommunications and finance. Headquartered in Richmond, Virginia, Amentra also has regional offices in Reston, VA and Charlotte, NC. Amentra’s web address is http://www.amentra.com.

2. Please describe the business and/or technical challenges you faced in this project.
The project had several significant business and technical challenges as outlined below:

Although these challenges were significant, they each also had reasonable solutions that could be addressed by a combination of proper project implementation and proper utilization of the JEMS stack. The challenge with the most far-reaching impact was the issue of the long-term implications of technology migration for the management of the IT department. This challenge alone had several major components:

– Any of the previous risks can negatively affect the total cost of ownership and return on investment – the gold standard of business success.

Although these risks are evident in every technology platform migration, they are rarely directly addressed and can often lead to the long-term failure or underperformance of a technology adoption effort and can poison the reputation of a product or technology solution.

3. What was the desired solution?
Amentra worked with the client to formulate a solution that involved two tightly integrated components: a traditional IT implementation with a focus on iterative implementation and heavy business involvement and a parallel mentoring approach that targeted developers, architects, IT support staff, and key business leaders.

Implementation and Architectural Approach

[ Note: At JBoss’s request, Amentra can describe every aspect of the business and technical solution (confidentiality agreements notwithstanding) in exhaustive detail. However, given that Amentra’s proposed innovation is its mentoring model for technology transfer and adoption of the JEMS stack, a brief overview of the technology solution will be provided for context while more attention is devoted to the mentoring aspect. ]

Amentra utilized its industry-leading expertise in J2EE implementation to help the client design a service-oriented architecture based on the JEMS stack. The architecture was specifically designed to provide scalable, reusable business and infrastructural services that would assist in the development of future applications. Amentra’s status as a JBoss partner also gave it additional insight into the future viability of various technology solutions on the JEMS stack, allowing further customization and refinement of the architecture.

A standard logical view of the architecture is provided in the JBoss World Innovation Award Submission
Mentoring Approach

The following section briefly explains Amentra’s mentoring approach as applied to this engagement and will provide a concrete case study of the mentoring process on a JEMS-centric project. The mentoring process is very flexible and based upon the level of the client staff’s experience and the client’s desired end result for mentoring, determined during the initial stages of the engagement. The process behind this methodology can be broken into several high-level steps:

Staff Skill Set Evaluation
Best Practices Opportunity Analysis
Mentoring Topic Customization
Delivery Process Planning
Periodic Review and Adjustment

Staff Skill Set Evaluation
At project inception, Amentra met individually with each member of the technical staff who would be developing or supporting the application in order to establish a basic understanding of the backgrounds and relevant experience of those to be mentored. Amentra focused on obtaining information such as the person’s job description, education and experience, as well as asking each person to complete a self-assessment on their specific business or technology skill sets. It was important that Amentra included support staff as well, as the platform would eventually impact every single person in the IT department.

The resulting feedback received from these individuals along with the end result expectations as described by client management was used to select not only the high-level topics to be covered during the initial mentoring sessions, but to calibrate the level of detail and focus that was targeted for specific topics. For example, even though the development staff all came from Visual Basic 6 and RPG development backgrounds, all of the team members had a solid basic understanding of SQL and basic relational database usage. Identifying this at project inception allowed Amentra to skip classroom training for that area and reallocate the time to discuss less well-understood areas like practical object-oriented design.

Amentra had similar, but more subtle, conversations with key business stakeholders. This allowed Amentra to help IT leadership craft an effective message that emphasized the platform’s strengths, but also communicated the platform’s limitations as well.

Best Practices Opportunity Analysis
As the initial skill set evaluation was concluding, Amentra conducted a review of the client’s business processes, requirements management approach, and/or software development lifecycle processes as appropriate to determine opportunities for refining, augmenting, or reducing process in order to become more consistent with current best practices for the new business and technology environment. In this case, the client had a fairly sophisticated business requirements gathering approach that would work well with the new technology platform. However, the development and testing approaches would benefit from different approaches that better aligned with modern J2EE development. Mentoring in these approaches was thus added to the mentoring plan.

In order to maximize relevancy, Amentra’s mentoring process has been designed to be extremely flexible in its ability to be incorporated within any lifecycle methodology. Amentra has its own iterative methodology for delivering turnkey projects and will utilize this process if the client has not yet developed a process. In this example, the client chose to be mentored on portions of these processes and incorporate these portions into their enterprise direction. Amentra has substantial experience in incorporating its mentoring strategy within very rigid environments for some of the largest companies in the world, including heavily regulated environments like the pharmaceutical, insurance, healthcare, and financial industries.

Mentoring Topic Customization
Using the findings from the staff skill set evaluation and the current status review, Amentra customized a mentoring approach for the client and the team being mentored. The approach highlighted multiple key business processes, technologies, and methodology topics as high-level subject categories for the mentoring effort as listed below. A non-exhaustive list of the mentoring topics covered includes:

Agile Methodology
Requirements Gathering
Test Plan Development
Unified Markup Language (UML)
Introduction to Object Oriented Programming
Basic Java/OO Programming, Section I
Basic Java/OO Programming, Section II
Advanced Java Programming
Java Server Faces (JSF)
Java Messaging Service (JMS)
The Enterprise Service Bus (ESB)
Database Design
Hibernate
Logging
JUnit and Grinder
Subversion (SVN) Configuration Management
Ant
Maven

Amentra’s extensive experience helped to focus on the most appropriate foundational mentoring topics for initial efforts in order to help prepare the client team for more detailed and nuanced mentoring later in the project.

Delivery Process Planning
Amentra worked with the client’s management team to coordinate the mentoring plan with the overall project plan for the engagement. Like the project plan, the mentoring plan had formal deliverables, timelines, and milestones. The mentoring plan was designed in compliance with Amentra’s following guidelines:

Delivery of mentoring topics were coordinated with the project schedule so that topics relevant to the current stage of a project are covered just prior to and during that stage whenever possible. These topics sometimes spanned different groups participating in an overall mentoring approach and were executed in parallel with these groups by different members of the Amentra consulting team.
Classroom training was always confined to a limited period of time as knowledge retention drops off sharply in long training sessions.
Mentoring material preparation time for extremely customized mentoring topics was considered. However, since Amentra has already created a significant library of mentoring presentations and material, additional preparation time was typically quite small.

Mentoring Delivery
Amentra then iteratively implemented the mentoring plan with the client. Initial iterations for each topic covered involved relatively short (1-4 hour), highly interactive classroom training sessions. This helped establish a baseline among the team for new topics and provided some structure for how the new technologies and skill sets could best be used. Most classroom training sessions had a corresponding set of “homework” assignments for the team to complete individually. This allowed the team to immediately reinforce the learning. Just as importantly, the assignments provided Amentra with immediate feedback on the amount of comprehension that occurred on an individual basis. In one or two cases, training sessions were repeated or extended based on the results of the assignments. In other cases, planned follow-on sessions were eliminated when the team demonstrated immediate understanding of the subjects.

The classroom training sessions were carefully scheduled to be executed immediately before a corresponding opportunity to use the knowledge in practice. One the classroom training established a baseline of comprehension, Amentra immediately targeted follow-on project tasks that helped ensure retention of the knowledge. As the staff attempted to apply their new knowledge to a project challenge, Amentra consultants worked with them individually at various points each day to ensure that they were progressing towards an effective solution, and shared additional, more refined techniques as the staff demonstrated increasing confidence and competence with their new skills. This carefully planned, one-on-one mentoring approach is unique to Amentra and has been critical in helping dozens of clients migrate from legacy platforms to more modern solutions.

It is worth reiterating that Amentra’s mentoring was not just applied to developers. Key business stakeholders and analysts were mentored in the software methodology and requirements gathering sessions. QA staff members were mentored in the test-related topics. Administrators and support staff were mentored in the introductory and administration-related topics. This holistic mentoring approach ensured that all stakeholders were up-to-speed in the new platform and techniques that were being adopted.

Periodic Review and Adjustment
The effectiveness and progress of the mentoring plan was periodically assessed and adjusted throughout the project as Amentra worked with the client to design and implement the application. This iterative approach to mentoring allowed for adjustments to be made as Amentra saw evidence of strengths and weaknesses in the new approaches, creating an optimal learning experience for the project team.

Summary
Amentra’s mentoring model ensured the long term success of the effort by addressing each of the following risks:

Potential Loss of Personnel/Business Knowledge – Amentra’s evangelization and individualized attention helped initially convince the client staff that they would continue to be vital members of the organization after the adoption of the new platform and would continue to be strong contributors to ongoing project success. As the mentoring process progressed, the staff became even more excited about the new skills that they were learning and applying on a daily basis. In fact, the IT department experienced no attrition at all among legacy developers during the project span.

Critical External Leverage – By training the client team in all aspects of product development and administration with the JEMS stack and other technologies, Amentra ensured that the client would be able to support and extend the application without any outside assistance.

Failure to Realize Productivity Gains – Amentra’s critical contribution to the long-term reduction in total cost of ownership was to mentor the team in optimal development practices using the JEMS stack and related technologies. This not only included detailed training in sophisticated development areas like remote application server debugging using JBoss-IDE, but also in software development best practices like designing for reuse, automating integration builds, and test-driven development.

Failure to Realize Integration Savings – Amentra’s vast experience in large-scale enterprise integration helped make this challenge simple for the client. Even before Amentra was formally engaged by the client, Amentra helped the client understand the attractiveness of an integration solution based on a reusable enterprise service bus. Once engaged, Amentra then provided critical mentoring that allowed the client to understand how to extend the integration implementations required for this project.

Increased Total Cost of Ownership – The shared knowledge provided by Amentra in each of the preceding bullets helped to ensure the smooth transition from the legacy technology platform to a JEMS-based platform and guaranteed that the client staff had sufficient in-house expertise to continue to deliver systems efficiently on the new platform.

Inappropriate Long-Term Expectation Management – As noted above, Amentra’s mentoring methodology has evolved over time to include informal mentoring of key business stakeholders specifically to ensure that expectations are properly managed.

Amentra’s innovative mentoring approach to project delivery and the client-consultant relationship has delivered initial project success on the JEMS stack for customers while ensuring their satisfaction with JBoss and Java for years to come.

4. Please describe your vendor selection process and why you chose JBoss Solutions in the end.
Amentra worked with the client to evaluate the JEMS stack along with several other commercial software vendors and several partial J2EE-based solutions (e.g., standalone portals, standalone servlet engines) for features, adoption costs, expected productivity, support capabilities, and licensing costs. JEMS was the clear winner in each of these categories.

5. What role did Red Hat and/or JBoss products play in the final solution?
At Amentra’s urging and with full client agreement, JEMS products played critical business and technical roles in the solution. JEMS products are used at every layer of the implementation, including:

Presentation Layer – JBoss Portal has provided the presentation infrastructure for the effort and has served as the interface into several of the reusable services designed for this effort (e.g., authentication/authorization, reporting).

Business Layer – JBoss Server has provided the central hub for the application and hosts the services that comprise the application.

Integration Layer – Hibernate has been used exclusively for all database integration and has drastically reduced the development time for this layer.

With Amentra’s encouragement, the client adopted Eclipse as the IDE of choice and leveraged the JBoss-IDE plug-in as well to help speed development.

Although the JEMS stack played an absolutely mission critical role in the technology stack, its most critical contribution was to allow the adoption of an enterprise-class, fully-supported J2EE solution at a price point that led to quick return on investment. Without this capability, the project might well have languished in the planning stage.

6. What was the overall impact of the project on your business? (e.g. improved ROI, increased competitive advantage, better time to market, etc.)
Amentra’s mentoring approach gave the client the confidence to include a significant amount of functionality in scope for the first release of the platform. This created several critical and immediate positive benefits for the business:

Reduction in Labor Costs – Within seconds of any student or employee arriving or departing any of the client’s branch locations, the system is notified and recalculates the appropriate labor staffing ratio based on regulations at the state, county, and municipality level. Management in the field is instantly alerted if staffing is too high and can react appropriately. Managers can then react appropriately and with iron confidence to minimize overstaffing. This significantly reduces labor costs, the largest single expense for the client, while maintaining excellent quality of service for customers.

Increased Regulatory Compliance – State, county, and municipal ratios are now automatically calculated based on centrally maintained information instead of being calculated manually at each branch location. This eliminates any chance of inadvertent non-compliance at the branches.

Greatly Increased Operational Visibility – For the first time, corporate management now has near-real-time reporting capabilities on attendance data. This allows for true auditing capabilities from the corporate office, increasing management efficiency in the field and ensuring that every location is meeting or exceeding all appropriate staffing regulations at all times. The use of JBoss Portal as a web interface and delivery method also allows district and regional managers to use the system for self-service reporting when traveling, a critical capability for an organization where some districts cover tens of thousands of square miles.

The savings and operational improvements noted above fully justified the implementation on their own. However, Amentra used their longstanding J2EE expertise to help the client design the system as an extensible, service-oriented platform that can quickly and inexpensively support additional capabilities in future versions such as:

Improved Strategic Reporting – Because of Amentra’s mentoring approach, the client now has the JBoss Portal expertise required to easily deploy existing reports to executive and field management through the JBoss Portal-based interface designed as part of this application. This will also allow the client to further leverage the common authentication/authorization service built during this effort. Further, strategic reports can now be updated on a daily basis instead of a weekly basis due to the ESB-based common data collection infrastructure (q.v. above).

Yield Management Analysis and Improved Pricing Models – Amentra helped the client design the business rules service in a manner that will also support rule-based pricing as part of a future effort. Utilizing more sophisticated pricing methods will allow the client to increase their revenue in the future without a corresponding increase in labor costs. The common data collection infrastructure is a necessary prerequisite for this capability as well, allowing for models that react instantly to changes in student attendance and staffing levels.

Centralization and Portal-Based Delivery of All Applications – The success of this JEMS-based rollout and the low associated development costs have made it likely that more of the applications that are currently executed at the branch will be centralized. This will eventually allow complete central data storage, reducing the computing needs at the branch level and eliminating the existing data protection needs at each branch.

Increasing Automation of Complex Business Processes – The client’s newly acquired ESB experience has enabled them to more aggressively target automation efforts that span systems. This has created a paradigm shift for the client that will likely support years of future projects that generate further incremental cost improvements. Detailed knowledge of existing systems and processes will be equally as important to the success of these efforts as ESB expertise, but Amentra’s ability to retool the development team with ESB skills has ensured that the system knowledge acquired over years of experience at the client has been preserved for the future.

Amentra’s expertise helped the client correctly design the initial services to readily support these future initiatives for very low effort. Amentra’s mentoring methodology ensured that the client developed their own in-house expertise to implement these initiatives with little or no outside support.

7. With the savings gained from implementing JEMS, how did you reallocate your cost savings within your company?
Confidentiality agreements with the client prevent Amentra from disclosing details of savings and expenditures at this time. However, some of the savings created by using JEMS was used to help implement additional services in the service-oriented architecture that will greatly lower future implementation costs for the client.

8. Please provide a technical description of implementation, including the size of deployment. (i.e. Hardware specs, applications, O/S, databases, etc.)
In order to fully leverage the client’s existing infrastructure standards and investments, the following hardware/software stack was used:

Presentation/Application Servers: JBoss Portal, JBoss Server, Hibernate, Windows 2003, on a cluster of 2-CPU Dell Enterprise-Class Servers

Enterprise Service Bus Servers: Cape Clear ESB, Windows 2003 on a cluster of 2-CPU Dell Enterprise-Class Servers

Business Rules Engine Server: Fair Isaac Blaze Advisor

Database Server: Microsoft SQL Server, Windows 2003 on a cluster of 2-CPU Dell Enterprise-Class Server

Business Intelligence Server: Information Builders WebFOCUS on existing hardware.

In order to meet the client’s needs, the application will support thousands of simultaneous users and hundreds of thousands of messages per day from over six hundred branch locations.

9. Did you leverage Red Hat support services, training, or consulting? If so, please describe your experience?
At Amentra’s suggestion, the client purchased JEMS support to guarantee support, warranties, and indemnification equivalent to that provided by a closed-source platform. Due to Amentra’s support and mentoring, the client has enjoyed the best possible experience with their support – they have not yet had cause to use it at all!

10. Advice to other companies considering JEMS.
The obvious licensing cost and standardization savings resulting from the adoption of professional-grade, open source platforms have traditionally been countered by the perceived difficulty in quickly retooling existing staff to effectively use these platforms. The combination of JBoss’s demonstrated commitment to platform excellence and the proven results of Amentra’s mentoring methodology for retooling legacy developers from COBOL, RPG, VB6 and dozens of other programming backgrounds have overcome this challenge and drastically lowered the entry cost for J2EE platform adoption.

Category: Business Process Auotmation

Winner:Helio/SK Telecom

Submitted by: Junwon Lee, Manager, Convergence and Internet R&D Center, SK Telecom

Industry: Telecommunications

Geography: Korea/USA

Overview
SK Telecom (www.sktelecom.com), Korea’s leading wireless telecommunications services provider, and Helio (www.helio.com), an advanced U.S. mobile service provider, were selected for their use of JBoss Enterprise Application Platform and JBoss jBPM to build and streamline its wireless data portal (WDP). After deploying Red Hat solutions, Helio quickly experienced cost reductions. Operational savings amounted to 50 percent monthly and the company experienced a 90 percent reduction in commercial software and licensing costs.

Please describe your company. (Number of employees, private/public, industry, etc.)

SK Telecom
SK Telecom is a mobile network operator in Korea and the leader in the Korean telecommunications market. SK Telecom has over 50 percent of the market share in Korea with nearly 20 million subscribers. The revenue is $11,368 million USD and net income is about 1,544 million USD. (based on 2006, 1USD = 937KRW)

Helio
Helio is a U.S.-based MVNO (Mobile Virtual Network Operator) and has about 600 employees.

Please describe the business and/or technical challenges you faced in this project.

Helio’s core mobile Internet infrastructure system, wireless data portal (WDP), serves as a mobile web portal, manages and delivers mobile content, and manages customer/device information. Its service was released to the market in May 2006. Simultaneously, WDP 1.0 was launched to provide wireless data service to subscribers. This system was implemented based on the Unix operating system WebLogic, Oracle, and commercial BPM software.

WDP 1.0 was not designed for U.S. businesses, so it was difficult for Helios to work with U.S.-based business partners, and a BPM tool in the system often malfunctioned, resulting in significant amount of data loss. Having been built on a commercial operating system, middleware, and application server, the maintenance, licensing, and hardware costs surged as the number of subscribers grew.

What was the desired solution?

The desired solution was to create a joint migration project, substituting WDP 1.0 with WDP 2.0, an open source-based platform, splitting resources and budget between SK Telecom and Helio. The migration would resolve the problems inherent with using using the previous system and to meet new requirements for working with U.S.-based providers. These changes would be reflected in the new system, WDP 2.0. Priorities included minimizing the system management cost, and offering better functionality, performance, and stability.

Please describe your vendor selection process and why you chose JBoss in the end.

The WDP 2.0 project had several missions to complete, which involved incorporating middleware, support, and flexibility.

When a runtime environment is constructed with JBoss, instead of being single-use, it can be paired with a JMX management tool, which saves on maintenance. Also, with open source, costs are low, but the technology stays competitive with higher-priced commercial solutions. Additionally, open source model development provides more secure and higher-quality code, and there is no vendor lock-in. JBoss also offers an impressive customer base and track record.

Describe the application you built using JBoss. What role did JBoss and/or JBoss products play in the final solution?

WDP 2.0 is comprised of four server groups: web, application, download, and database. JBoss Web Server (WS), JBoss Application Server (AS), and jBPM modules are installed on each server in accordance with its usage. The new WDP 2.0 platform has various types of business processes that administrators have to perform or confirm. Business Process Management (BPM) enables administrators to monitor and control the “to-do” list in an easy and flexible way. Hence, every request submitted by content providers can be completed very quickly, turning a very complex process into a very simple one via the Business Design Tool.

What value did you gain from implementing JBoss solutions and how did this impact your business? (e.g. improved ROI, increased competitive advantage, better time to market, etc.)

Helio reduced system maintenance costs, saved approximately 65 percent in hardware costs, saved 90 percent in purchasing commercial software and paying maintenance fees, and saved 80 percent for systems operation and maintenance in comparison to its previous solution. Additionally, Helio saved approximately $25,000 per month on service operational costs.

With the wireless data portal platform based on open source, the initial system construction cost for SK Telecom will be reduced significantly when launching services in an international project. Consequently, this means that SK Telecom can have more opportunities for global business, including in China and Southeast Asia.

Please provide a technical description of implementation, including the size of deployment. (I.e. Hardware specs, applications, O/S, databases, etc.)

WDP 2.0 platform enables enhanced support for new target deployment platforms. With the added flexible support requirements & features, installation for alternate operating systems will be made possible, including the various flavors of Linux available in the marketplace. Additionally, the targeted runtime environment will be supported on open source platforms and systems, and providing support for other J2EE application servers adds enhanced flexibility in installation and production deployment scenarios. Adding the JBoss Application Server (AS) to the mix as a target deployment platform also helps to minimize the licensing costs and overall cost per deployment. WDP 2.0 is composed of two web servers, two web application servers, two download servers, and three Database servers.

Did you leverage JBoss support services, training, or consulting? If so, please describe your experience?

We utilized consulting for the migration from WebLogic to JBoss in order to help predict any problems that might result from the switchover, as well as what the managing strategy should be for the revised project plan. The results of the consultation were effective numbers for the necessary amount of man power and the timeframe for the switch, which helped contribute to a successful completion of the project.

The main use of consulting was concentrated on the BPM tool conversion, mostly on the functional differences between the commercial BPM and jBPM. Thanks in large part to the advice gained, a successful conversion to jBPM was realized and has since been successfully managed.

For the successful WDP project, we had our developers receive JBoss training. After the training, they have, in turn, trained the other developers. This series of education induced them to develop projects with a clear understanding of JBoss technologies and methodologies.

Do you have advice for other companies facing a similar business challenge?

With JBoss’ open source solutions, we were able to adopt advanced technology that was available for a fraction of the price of our previous solution. We were hesitant to adopt an open source solution at first, but encourage other to have confidence in open source technology because there are excellent support options and resources available. We have had no problems with our JBoss solution and are producing higher-quality service for a lower cost.

Category: Service-Oriented Architecture (SOA) Implementation
Winner:Vivat (client = US Trust)
Submitted by: Max Yankelevich, Chief Architect
Industry: Vivat = Technology Partner / US Trust = Government, Financial
Geography: Bensalem, PA

Overview
Vivat (www.vivatconsulting.com), a professional services organization focused on Service-Oriented Architecture (SOA) and Enterprise Application Integration, was selected based on its work with US Trust, one of the oldest private banking firms in the U.S. Vivat successfully moved US Trust’s main revenue stream, the Client Fee Calculator, from an antiquated legacy system to a new SOA system built on JBoss technologies, including JBoss Rules and JBoss Application Server. The solution is expected to bring in an additional $20 million over the next five years as well as maximize customer profits, thanks to newly transparent and manageable fee rules.

Please describe your company. (Number of employees, private/public, industry, etc.)

Vivat is a boutique professional services organization, focused on Enterprise Application Integration, Service Oriented Architecture (SOA) and Technical Architecture of Enterprise level components, that delivers superior service and cost saving to its clients by coupling exceptional talents with deep understanding of the industry. With over 100 employees, Vivat is one of the largest “small” and perhaps most successful SOA firms on the east coast.

Please describe the business and/or technical challenges you faced in this project.

Our client, one of the oldest Wealth Management and Private Banking firms in the US, was looking to re-engineer its key, proprietary engine, the Client Fees Calculator, which was responsible for the entire firm’s revenue stream. Highly customized and complex in nature, Fees Calculator was created over 20 years ago as a large set of Mainframe programs. Fees were calculated, in batch, on monthly or even yearly bases. Revenue was lost due to the lack of understanding how the code base actually worked under current market conditions, since most of the original Assembly and COBOL programmers have left the company or retired. The firm also had to deal with numerous industry compliance issues due to the above mentioned challenges.

What was the desired solution?

Business stakeholders were looking to move to a more real time fee schedules, as well as the ability to change and model fee rules on per customer bases and to use in financial projections. Technology stakeholders were concerned with making the Fee Calculator more maintainable and scalable, as well as migrating it to a distributed, commodity, cost effective technology stack from the expensive Mainframe platform.

Please describe your vendor selection process and why you choose JBoss in the end.

Professional Open Source software was the focus of the selection process, as we were looking to reduce the project’s overall budget by avoiding software license expenses. We were also looking for modular and flexible component set, with good industry acceptance. JEMS product suite provided the most complete offering for the project with the best support quality and options.

Describe the application you built using JBoss. What role did JBoss and/or JBoss products play in the final solution?

The new Enterprise Fee Calculation Service was built in strict compliance with Service Oriented and Event Drive Architecture principles. JEMS product suite provided all of the necessary building blocks to help in successful implementation of this project. Red Hat Linux OS and JBoss Application Server provided the core run-time platform for all the layers of the application. Financial transactions, such as trades, bank deposits/withdrawals and loan payments, were received in real time through the Messaging layer (JBoss Messaging). The receipt of a financial transaction would kick off a Fee Calculation Business Process (jBPM). Fees were recalculated by the Rules Engine layer (Drools/JBoss Rules), which housed over 5,000 core business rules. The new fee amounts and details were stored in the relational database (MySQL) using Object/Relational layer (Hibernate).

What value did you gain from implementing JBoss solutions and how did this impact your business? (e.g. improved ROI, increased competitive advantage, better time to market, etc.)

The estimated Return on Investment (ROI) was calculated to be $20M over next 5 years, as this was the amount that the firm undercharged their clients by every year. Also, profits were maximized on per client basis because the fee calculation rules were transparent and manageable, hence allowing for creative deals to be struck on the fly.

Please provide a technical description of implementation, including the size of deployment. (I.e. Hardware specs, applications, O/S, databases, etc.)

Enterprise Fees Calculation Service was deployed on top of Red Hat Linux / JBoss Application Server cluster spanning across 2 HP DL360’s. The application stack included Drools/JBoss Rules engine, jBPM, Spring and Hibernate. The database cluster was running MySQL RDBMS on top 2 HP DL380 boxes with GFS deployed over SAN.

Did you leverage JBoss support services, training, or consulting? If so, please describe your experience?

JBoss Enterprise support contract was money well spent! We fully leveraged JBoss support organization and tools, got very positive results, quick problem turn around and concise answers.

Do you have advice for other companies facing a similar business challenge?

Practical expertise in implementing SOA solutions is key when undertaking and Enterprise transforming initiatives. Don’t be fooled by SOA software solutions as being the answer – expertise of bringing process, software and people together is what gets rusults!

Category: Increased ROI
Winner: Alintec
Industry: Tech and Government
Geography: Milan, Italy

Overview
Alintec (www.alintec.it) (previously Politecnico Innovazione), a non-profit consortium promoting technological innovation within SMEs and public institutions and technology transfer between academia and industry in Europe, was selected following the high return on investment (ROI) it achieved after the implementation of multiple JBoss technologies within its Internet-based Library Management System (LMS) for the Province of Bergamo. The JBoss-based solution increased library loan rates by more than 25 percent and the frequency of inter-library loans by over 30 percent. Overall, it has positively impacted the end-user experience in terms of speed and ease-of-use.

Please describe your company. (Number of employees, private/public, industry, etc.)

Alintec is a non-profit consortium, whose aim is the promotion of technological innovation within SMEs, public institutions and the cooperation and technology transfer between University and Industry in the European dimension. We support research and technological development (RTD) activities by:

• Acting as Project Manager providing key competences in the development of innovative projects. Coordinating multidisciplinary activities and embody the only company’s referent making cooperation with University easier.
• Evaluating technological solutions and the main strategies that should be adopted.
• Identifying the company’s needs for new technology in order to develop new products, to start up new productive processes and to improve the company’s business management.
• Developing applied research projects defining how to exploit the new technology as well as its impact on the enterprise or public administration.
• Technology transfer concerning information technology and its management.
• Analysis and support for the use of innovative ICT technology to make strategic changes.
• Promoting the cooperation between University and Industry Providing responsive and professional services to find, develop, apply and improve innovative solutions for small and medium enterprises, private companies and public bodies.

Coordinated by experienced faculty members, young engineers are encouraged to take part on the projects as needed and as motivated by their specialization.

Please describe the business and/or technical challenges you faced in this project.

The project goal was to design, develop and implement an internet based Library Management System (LMS) to integrate the 230 libraries of the Bergamo province. It was presented to us as a challenge, literally: commercial companies had tried to implement a solution to the needs of the library leading, operators, end users and public administration. They failed: the proposed solutions resulted too slow to accommodate all concurrent users, unmanageable, inflexible, buggy, crash-prone and the constant bug fixing had painfully slow resolution times and was very expensive. The original contractor preferred to back out from the support agreement; so the public administration came to university for advice, got it from our analysts team, was satisfied and so finally proposed us to build a new implementation using the technology and developing processes we had recommended them. The main challenge was quickly identified as the joined need for fast searches, changing business rules, and very high concurrency. Also, W3C WAI-AAA compliance was set as a requirement. The main business processes to support:

• Cataloguing
• Acquisitions
• Circulation
• Inter-Library Loan (ILL)
• Document reservation
• Web OPAC (Open Public Access Catalogue):
  • Public search engine
  • Open reservation
• Reporting The business numbers:
  • 230 public libraries
  • 500 operators (librarians and cataloguers)
  • 200,000 final users (citizens with library card)
  • 1.6 million loans by year (maximum 10,000 loans per day)
  • 4 million documents

What was the desired solution?

To simplify IT management and maintenance costs, they didn’t want to distribute software clients to all machines, to avoid client software management. A browser-based solution could meet this requirement, but they would prefer not to change radically the rich client user interfaces they were accustomed to: the interface should be similar to past solutions to avoid extensive training of all operators. The solution needs full web2.0 technology support, even library operators and cataloguers don’t want to install/maintain client software other than their browser. This is also a political need, with a higher usability more people will be using the system and the public bodies who endorsed the system will get credit for a good choice. In order to provide a good user experience a responsive system is a requirement: they wanted a fast system, and a lot of full text searching. They preferred to invest in fast and scalable software than to buy expensive hardware, as the Italian government endorses a software reusability program. During the analysis phase we were not sure we could provide the needed performance, so we were looking for a scalable architecture to eventually have the possibility to add hardware. We needed a flexible architecture to accommodate all requirement changes and to be able to provide the agility needed for new features and fixes, also speed and processing efficiency have always been a primary concern.

Please describe your vendor selection process and why you choose JBoss in the end.

We wanted to develop with open source components, so we began looking at JBoss, Spring and Struts. At the time, in May 2006, JBoss was providing an almost-full compliance with JSF and EJB 3.0 as JBoss RC9 was released. Seam looked very promising and innovative compared to other competitors and was backed by a team we trusted, as we already had had some experience with JBoss’s connection pool and hibernate on previous projects. As a research centre linked to University we were particularly interested in trying out bleeding edge technologies. Also JBPM and AOP were looking like the solutions to our flexibility needs, and the Seam recommended patterns found in the first examples released in those times looked very promising. We didn’t know yet whether we would need much more features such as web services and schedulers, but it was nice to know that they would be available when needed. After some preliminary testing the choice was easy.

Describe the application you built using JBoss. What role did JBoss and/or JBoss products play in the final solution?

We configured the complete stack from the hardware configuration, operating system up to the java enterprise ear. We use Fedora as operating system, Apache httpd to serve some static content and finally JBoss to run the application and to serve all dynamic content. The ear contains much more technology from JBoss:

• Seam: the whole application is extensively based on Seam, we are using it as core technology from first beta releases; we use conversations, Seam JSF tags, internationalization, iText reports, email support, remoting and jboss-el capabilities.
• RichFaces: obviously, to implement web2.0 rich clients we used many RichFaces JSF components.
• Hibernate: the complete persistence layer is built on hibernate 3.2, mostly through ejb3 persistence API, sometimes using hibernate annotations. Hibernate Search is a core component; when the beta appeared on the hibernate website we immediately began testing it. Unfortunately the timing was wrong and it was still too “on-the-edge”, so we ended up using a mix of direct Lucene “low level” and Hibernate Search for other functions. We had to code direct index access to tweak the most complex queries and be able to search on structured objects, as these features are now available through Hibernate but were not at the time of release. You can take a look at the system at: http://opac.provincia.bergamo.it.

What value did you gain from implementing JBoss solutions and how did this impact your business? (e.g. improved ROI, increased competitive advantage, better time to market, etc.)

We began developing in October 2006 and released the final version in May 2007; now in October 2007 the customer has had some time to verify the value of the new product; we can summarize their satisfaction in the next points:

• Fast real time information sharing between all business units: some libraries are in remote mountain districts but are now served as well as in the biggest cities.
• Substantial ICT maintenance cost reduction. Better “time to user” in assistance and bug fixing (As it is a public service we wouldn’t call it “time to market”).
• Library loans increased of +25%; we believe this to be a good index to evaluate the final users satisfaction in the library system’s core mission, and a benchmark for usability.
• Inter-Library-Loans increased of +30%; This means the different libraries use available books more efficiently, reducing book-buying costs and final user’s wait queue.
• 24/7 service.

It is difficult to compare other indexes such as delivery delays to the older system as information is incomplete on the previous system, but even so the public administration and library operators agree that general efficiency has greatly improved. They are now so satisfied that they asked us to develop many more features; the JBoss technology we are using empowers us to provide the new features and functionalities at a competitive budget comparing to other solutions. Seam helped to integrate all different technologies in a clean design so that it is now easy to extend the current features: after having overcome the difficulties in using a young technology, Seam left us now in a win-win situation.

Please provide a technical description of implementation, including the size of deployment. (I.e. Hardware specs, applications, O/S, databases, etc.)

The service is made up of two dedicated servers, an application server and a database server. Backup services and network management are provided by the hosting company. Application Server: 1 HP DL360 G5 (2x Xeon 5160 with 8GB RAM) OS: Fedora Linux 6 / 64bit JBoss 4.05 (now upgrading to 4.2.2) Database Server: 1 HP DL380/1 (1x Xeon 2.8 GHz with 2GB RAM) OS: Microsoft SQL Server 2000 on Windows Server 2003 A single JBoss on a single JVM (v.6u3) is currently enough to satisfy the performance demand; It supports clustering so we are going to add a second machine to support failover, but the provided uptime is excellent even now. A custom JSF component has been built to page through search results, linked to a special type of beans to control all our different kinds of searches: implementing new query flavors became flexible, independent from the technology used to retrieve the data: by using plain Hibernate, Hibernate Search, direct Lucene, native SQL, by external web services, by remote legacy or standard library protocols. We are considering the possibility to give this code back to the community as it demonstrates very fast Hibernate queries but this will need some code polishing; a full-text query is done on average in less than 16ms, testing under high load. A custom high speed Lucene indexer was built, we’d love to give more details for this too but think it could be far too complex for a generic simple application; we are now able to index 4 Million structured documents (each being represented by a dozen of linked entities) in less than 20 minutes. During day CPU’s usage doesn’t go higher than 30%; at night some background jobs are started by quartz to build usage reports, DB cleanup, Lucene’s indexes optimization. The back office management in the library system is very extended, there are currently 15 different user roles defined to protect 330+ views, only 10 of these are public, brought to life by 760+ beans.

Did you leverage JBoss support services, training, or consulting? If so, please describe your experience?

During first beta releases of Seam nobody knew how to develop with it; also books were not yet available: best practices were unknown, nobody knew how to get things working. So we had to learn how to interact with the JBoss Community, asking for help, submitting bug reports, sharing the little we had learnt. This was a new experience for all of us, but revealed very interesting: we finally understood the full power of open source communities, discovering that the people that works behind these technologies were incredibly helpful and skilled. Also when no other seemed to know the answer, we got help directly from Gavin King and Pete Muir, who resolved our issues very fast. In the past we were not quite interested in the possibility to give something back to the community, now is this same possibility regarded as a very high achievement. We found almost all knowledge we needed to develop this system on jboss.org’s wiki pages, documentation or forums, or looking to examples and to the source code of libraries themselves.

Do you have advice for other companies facing a similar business challenge?

We will definitely use the same technologies again, both for simpler projects and for larger systems. We had some difficulties in the first months because of lack of competence, so we would recommend other companies to make use from the JBoss support services as in home training resulted very costly both in terms of time and of developer’s exhaustion; also we still would have many questions even now it is working quite well. We had to use a legacy database with very new technologies because of client’s desire; I would advice not to mix old and new and say a firm “No” to the client next time, as it is difficult to find other people with the same configuration to learn from and often these combinations are not tested as well; we are actually going to try java DB just to see how it performs, and eventually switch database technology. We would also recommend to participate in the communities both by forums and by code, as this revealed a very useful, clarifying and a fast road to solutions. Now that we have some skilled developers we are successfully using JBoss’s technologies on all starting projects.

2008 JBOSS INNOVATOR OF THE YEAR

Category: Ecosystem
Winner: Rivet Logic Corporation
Submitted by: Mike Vertal, CEO/President/Owner
Industry: Rivet = Technology Partner / Kaplan = Education
Geography: Reston, VA

Overview
Rivet Logic, a provider of professional services focused on open source solutions, and Kaplan Test Prep and Admissions, a premier provider of educational and career services for individuals, schools and businesses, were selected for their use of Alfresco’s content management platform in support of the upgrade of Kaplan’s online presence from a legacy system to a JBoss-centric solution. The teams employed JBoss Seam (and Facelets), JBoss Application Server, Hibernate and jBPM to create a next-generation platform for the www.kaptest.com site that can deliver personalized applications and dynamic, targeted content. The results include a 26x performance improvement over the legacy content authoring/delivery system, much faster page load times and a “fresher” web presence for Kaplan Test Prep and Admissions.

Please describe your company. (Number of employees, private/public, industry, etc.)

This is a joint submission by the following three companies:
Kaplan Test Prep and Admissions (KTPA) Nearly 70 years ago, Kaplan pioneered the test prep industry. Today, as a division of The Washington Post Company (NYSE:WPO), Kaplan has become the leading educational services company in the world—helping more than 3 million individuals achieve their educational and career goals through programs ranging from high school and college admissions consulting to graduate school, professional licensing, and English language training. Kaptest.com is the company’s portal to these programs. The site is divided into unique communities where customers can find programs, read articles, access special events, tools, and more—all geared towards their specific goals and interests.
Alfresco Software Ltd. Alfresco is the first and leading open source alternative for enterprise content management. It is the first company to bring the power of open source to the enterprise content management market, enabling unprecedented scale and a much lower total cost of ownership than proprietary systems.
Rivet Logic Corporation – Systems Integrator of Open Source Software; Red Hat Advanced Business Partner; Alfresco Gold Partner and North American “Partner of the Year” Rivet Logic provides professional open source services and solutions that help organizations engage with customers, improve collaboration, and streamline operations. We offers a full suite of JBoss professional services including deployment, customization, and integration — enabling clients to fully leverage the power of the world’s leading open source enterprise middleware stack. With complementary expertise in the Alfresco content management platform, Rivet Logic offers integrated content-rich and SOA-enabled solutions that power a new generation of interactive web properties, enterprise intranet applications, and collaborative Web 2.0 communities.

Please describe the business and/or technical challenges you faced in this project.

KTPA’s primary web presence for its potential and actual customers is www.kaptest.com, which includes 10+ domains, 14,000+ unique pages, and 250,000+ contens, and a personalized experience for hundreds of thousands of students. The legacy system challenged business operations in terms of time, resources, usability, and performance.

• Time: The legacy system sometimes required a 2 week publishing cycle because of technical issues.
• Resources: The legacy system required IT involvement to support both web content authoring, and publishing content to the web content delivery system.
• Usability: The legacy system had an editorial authoring environment that was not user friendly.
• Performance: The legacy system was not performing well as traffic load increased, and scaling it out required additional high cost licenses.

What was the desired solution?

KTPA required a solution that:

1. Supported publishing on demand.
2. Required no IT involvement during content authoring, content review/ approval/workflow, and publishing to the web.
3. Supported robust enterprise architectural principles, such as clean separation of content from presentation, and web content management from web application development.
4. Had an intuitive and user-friendly authoring environment.
5. Incorporated a fast, high performance application stack and content delivery framework. Further, it should be scalable, stable, enterprise level, inexpensive and be based on Java.
6. Allowed KTPA to pay for support, not for licenses.
7. Enabled reuse of components, as supported by Seam, Facelets, and JSF.
8. Support an open source, Java based CMS (Alfresco).
9. Quality support should be available and good community background.
10. Should use open standards (not proprietary).
11. Support a designer-friendly framework for the presentation and a pluggable, middle tier framework. Templating, re-use of content, and ease of design/development were top priorities for the presentation framework.
12. And session management was a key requirement to the project. Components that can be controlled based on the scope of the session was a key technical requirement for the platform we were going to choose.

Please describe your vendor selection process and why you chose JBoss in the end.

One of the primary reasons JBoss was selected over other open source frameworks/stacks was KTPA’s past experience with Red Hat/JBoss support, which in our experience was always outstanding and extremely responsive. In addition, the following attributes of the JBoss/Seam/Facelets stack were critically important:

1. Cutting edge technology, allowing us to start innovating with our web presence.
2. Very high performance.
3. Open architecture, Open source
4. Easy integration with the Alfresco web content management platform.

Furthermore, the JBoss Seam + Facelets combination helped our content delivery system framework to achieve a true, enterprise-class modern web presence and platform.

Describe the application you built using JBoss. What role did JBoss and/or JBoss products play in the final solution?

The entire kaptest.com infrastructure is now running on what we call the Enteprise Content Authoring and Delivery System (ECADS), which was built by Rivet Logic and KTPA using JBoss AS, Seam, Facelets, Hibernate, jBPM, and Alfresco. ECADS is a system for authoring and publishing of web content for multiple, commonly hosted, websites. ECADS leverages the robust Alfresco XForms-based, multi-sandbox, layered versioning, content authoring system as a foundation for web content authoring (along with jBPM for workflow); and combines that with JBoss Seam’s powerful bijection, fine-grained scope management, and JSF extensions for content delivery. The content delivery framework portion of ECADS was built on JBoss Seam and EJB 3.0 with Hibernate-based JPA. The bulk of the delivery framework was built using Seam-wired POJO-based services with various scoping to accommodate the different life-cycle requirements. Lower layer services provide a foundation and an abstraction for the upper layers and is application scoped; where as higher layers that present data to the user were session scoped Seam Listeners answering directly to JSF components. The layers comprising the content delivery framework are:

1. WCM Layer: Abstracts the run-time instance of the Web Content Management System.
2. XML Layer: Handles the parsing and merging of the XML descriptors.
3. Accessor Layer: Domain specific layer that shares the context (from a terminology perspective) of the content-rich application (website).
4. Presentation Layer: Renders the final result to the user.
5. URI Transformation Layer: Transforms inbound/outbound URIs based on a pluggable transformer pipeline The presentation layer is especially interesting as it makes heavy use of Seam bijection to render the final page to the end user. And Alfresco was extended to support JBoss Seam for both preview and final publishing of websites.

What value did you gain from implementing JBoss solutions and how did this impact your business? (e.g. improved ROI, increased competitive advantage, better time to market, etc.)

The new kaptest.com was launched in early November 2007, and since then the value gained has been manifold:

1. Streamlined business operations and improved productivity. The KTPA editorial team has complete control over the content authoring and publishing process, with no involvement from IT required.
2. Better customer experience through much higher performance. We are seeing up to 26x performance improvement with the JBoss-based web application/content delivery system when compared to the legacy system. Specifically, current kaptest.com page load times are typically less than 1 second even under high load (which is remarkable for a dynamic web site!), whereas the legacy system took up to 13 seconds for a page load. Browse www.kaptest.com to see a high-performance JBoss/Seam application in action.
3. Improved productivity and a “fresher” web presence through on-demand web publishing: KTPA editors can now author and publish new content to the web within 8 minutes, contrasted with days and weeks with the legacy system.
4. Much higher ROI because of zero licensing costs. KTPA resources are now focused on support and development of innovation, instead of software licenses.

Please provide a technical description of implementation, including the size of deployment. (I.e. Hardware specs, applications, O/S, databases, etc.)

The production environment for kaptest.com is a multi-tier, multi-node system in a high-availability configuration. Key specifications include: Hardware nodes include: 8GB RAM 64bit CPUs Software installed includes: Red had Linux 2.6.9 EL Apache2 Mod JK JBoss AS 4.0.5 Java 5 – JRockit on 64 bit hw Facelets JSF RI 1.2 JBoss Seam 1.2.1 EJB 3(JBoss Hibernate) EHCache standard OS libraries like Apache Commons Alfreco WCM v2.0.1 Oracle 10g.

Did you leverage JBoss support services, training, or consulting? If so, please describe your experience?

For the most part, the JBoss stack and Seam framework were very stable. We utilized JBoss support for a handful of issues, mainly related to advanced, cutting edge features in Seam. JBoss support was very responsive, providing bug fixes and patches to resolve issues. JBoss (and Alfresco) consulting and development was provided by Rivet Logic.

Do you have advice for other companies facing a similar business challenge?

JBoss Seam and Alfresco WCM is a wonderful platform for enterprise-grade, high-performance, rich content delivery for next generation web sites.

Category: User Experience
Winner: Daiwa Securities America
Submitted by: Steve Dunstan, Vice President / Enterprise Systems Architect
Industry: Financial services
Geography: New York, NY

Overview
Migration of 120 separate applications running to JBoss Portal using a series of unified templates. This enabled rapid growth, increased productivity, faster service, and hundreds of thousand of dollars in cost-savings.

Please describe your company. (Number of employees, private/public, industry, etc.)

Daiwa Securities America is a wholly-owned subsidiary of Daiwa Securities Co. Ltd. of Japan. The subsidiary focuses on sales and trading of Japanese and U.S. equities and fixed income instruments, financial futures, and investment banking, including mergers and acquisitions and structured finance. For the year ending March 31, 2007, Daiwa Securities America had ¥14,684,000,000 (134 billion dollars) in net operating revenues.

Please describe the business and/or technical challenges you faced in this project.

In 2006, we successfully migrated one section of our prior portal to JBoss Portal. The section we migrated contained our Compliance Dashboard, a portal that ensures new financial and securities compliance regulations remain updated and implemented across our company. We initiated the migration because of the serious problems we were having with our prior vendor—not just technical, but organizational, and with tech support in particular. Because we are small and agile, we were more bleeding edge than their other customers; we had stressed the prior vendor’s platform to the breaking point. The prior vendor’s support team couldn’t answer any of our questions and were unable to support their own product. In addition, the slow speed and unreliability of the prior vendor’s portal was becoming an issue for our users, and the complexity of porting new applications was an ongoing problem for our IT department.

Based on the ease of the Compliance Dashboard migration and the resulting stability we experienced, we decided to migrate the remainder of our applications onto a new JBoss-based portal. Because we are on the leading edge of technology, we also needed a solution that we could support ourselves. This gave an open source solution a natural competitive advantage.

What was the desired solution?

We required a high-performance solution that was stable, fast, and flexible. Our IT department needed to be able to roll out new applications quickly and insert them on-the-fly. We also needed to gain more control over the vast number of applications that were running on the portal because supporting them was difficult and time consuming.

Please describe your vendor selection process and why you chose JBoss in the end.

When we had the problems with our prior vendor, we brought in a large, well known IT vendor for an interview. From their very high-level PowerPoint presentation, it became apparent that our agile corporate culture was not in line with theirs. We also evaluated another vendor’s product, but it was overly complicated.

On the other hand, our previous experience with JBoss products and technical support had been excellent. We had purchased a support subscription for the Compliance Dashboard and used it during our JBoss migration. When we had problems with the applications we were building, we were able to leverage JBoss support, sometimes communicating directly with the engineers who actually coded the program. The high quality of first-line support JBoss provides impressed us. We don’t have to go through a bunch of call center people to get answers to a problem.

Describe the application you built using JBoss. What role did JBoss and/or JBoss products play in the final solution?

Our portal had about 120 applications built on top of it, and we weren’t sure how we were going to perform the migration. We completed a broad analysis of the applications, looked at what they did and what they were used for, and found that they all did pretty much the same thing. They were mostly database-driven applications with minor differences in the inputs and outputs (commissions, broker dealer set-up, client set-up, etc.).

We used JBoss Portal to front end the database. Then, we designed an idealized template that described our applications in terms of how they “mine” that database, and ran the Velocity templating engine to regenerate all of our applications on the new JBoss Portal. Once we decided how to do this, it only took us three months from development to production. To complete the migration took one year. We had four people working on it: one person worked on the bulk of the templates; three developers handled the others that were more complicated or needed different business rules.

What value did you gain from implementing JBoss solutions and how did this impact your business? (e.g. improved ROI, increased competitive advantage, better time to market, etc.)

• From the user’s perspective, the biggest gains were speed, reliability, and ease of use. Since all of the applications are now from a common source, the method of operation and presentation of data is consistent from screen to screen. As a result, we’ve reduced training cycles for employees.
• From the developers’ point of view, we can create new applications in far less time. The compile time for an application dropped from five minutes in the prior vendor’s portal to less than a second in JBoss. In addition, the modularity of JBoss allows us to plug new pages into the portal easily. With our prior vendor, we either had to roll the entire portal (which took over an hour), or use WSRP (Web Services for Remote Portlets), which had lots of inherent problems. With JBoss, we’ve experienced much shorter development cycles, which means we can be more innovative in our approaches.
• The JBoss solution is also easier to maintain. With our prior vendor’s product, we had some difficult, opaque technical problems that we were never able to get resolved. Now, when we have a really tough problem, we look at the JBoss source code to help us debug it.
• Because we’re only running one program, the applications are also easier to maintain. When we find a bug, we re-roll the applications and the bug fix propagates to all of the pages.
• The system is far more reliable, as well. During the last year, we have had no unscheduled down time.
• Our IT team’s credibility with management has improved significantly. Since our applications no longer crash, management is now comfortable coming to us with their problems and required improvements. Turnaround time for developing new applications they request is only a few weeks.
• JBoss has provided us with the scalability we require. Since the time we switched over to JBoss, the number of applications running has increased from 120 to 170, and continues to grow at 40% a year.
• In strictly monetary terms, we’re saving $90,000 by eliminating the prior vendor’s support costs.

Please provide a technical description of implementation, including the size of deployment. (I.e. Hardware specs, applications, O/S, databases, etc.)

Our back-end database is Sybase running on a Solaris server. For development, we are using a single four-CPU server that is front ended by Apache on a single CPU Solaris box. For production, we have a Windows 2003-based server online, with a hot backup (the total number of servers is doubled because of our disaster recovery setup).

Did you leverage JBoss support services, training, or consulting? If so, please describe your experience?

Our key people took the JBoss for Administrators class 18 months ago. We currently have a Premium Subscription to the JBoss Portal Platform and are very happy with the high level of support we receive. During the migration process, we also worked with a JBoss consultant. We had some tough integration problems related to security, so we asked for help from JBoss. They got the job done.

Do you have advice for other companies facing a similar business challenge?

If you’re going to be a fast, nimble organization, you need to consider open source options. The ability to self-support yourself means that your down times will be minimized. It also means that you can develop cutting-edge applications by exploiting the openness of the program.

Download this video: [Ogg Theora]

Category: Migration
Winner: Sakonnet
Submitted by: David Renton, Vice President, Product Delivery at Sakonnet
Industry: Oil and Gas
Geography: New York, London, Rio de Janeiro

Overview
Used JBoss Application Server and Red Hat Enterprise Linux to develop a scalable, performance-driving, well-supported solution that provided ease-of-use, cost savings, and transparency for Sakonnet’s energy-trading solutions. With JBoss Application Server, Sakonnet discovered a solution that combined low-cost, quality software and reliable support to fit into its Software-as-a-Service (SaaS) strategy of frequent application upgrades and top-notch support.

Please describe your company. (Number of employees, private/public, industry, etc.)

Sakonnet Technology, dedicated to excellence in trading solutions for wholesale energy markets, is a private company that was started by JPMorgan technologists in 1999. The company focuses on energy trading and risk management systems under the Software as a Service (SaaS) model for utilities, banks, hedge funds, and commercial and industrial companies that are heavy consumers of power./p>

Sakonnet’s main offering is the energy-trading application, Xenon®. Xenon is the ideal platform for running a profitable, tightly-controlled trading operation. It allows clients to set their processes and controls at every stage, from pre-trade analysis to final settlement. The application covers trades in power, natural gas, crude oil, refined products, coal, and emissions credits.

Please describe the business and/or technical challenges you faced in this project.

The challenge for companies trading in energy markets is handling the vast quantity of data that arises from transactions combined with traders’ reporting needs. There is a tension between the amount of data and the near realtime speed with which traders need consolidated reports to take further actions in the markets. A single trade in the power market may require the system to capture a distinct price and volume for every 15-minute period for the next year, and then turn around and report on position and risk in a matter of seconds. Timely reporting and suitable follow-on actions are a key difference in making profit and avoiding loss in fast-moving markets.

What was the desired solution?

As we developed the newest version of our application, Xenon 5.1, we were focused on scalability and performance and needed a solution that could help deliver these benefits. Originally, the application ran on Weblogic technology. With the vast amount of data processing involved in energy trading, we required a large number of application servers and suffered heavy licensing fees associated with our previous Weblogic solution.

Please describe your vendor selection process and why you chose JBoss in the end.

When looking for a solution that would fit our scalability, performance-enhancing and cost-effective requirements, we looked toward open source solutions because we knew we could find the reliability and low cost that we desired. We decided to migrate to JBoss technology because we saw a solution that combined the transparency, which would provide us with top-level operational availability, and the same functionality from a licensing perspective, but with a much lowered cost. The quality of the software, the support available for the software, the performance data, and the ease of deployment were also key factors in our decision to migrate from Weblogic to JBoss.

Describe the application you built using JBoss. What role did JBoss and/or JBoss products play in the final solution?

Once we migrated to JBoss Application Server in 2002, we added compute server clusters, today known as grid computing, to our n-tier J2EE application for better scalability, while leveraging our existing infrastructure. We paid one licensing fee and were able to utilize JBoss solutions on scores of machines, avoiding a per-CPU cost model.

What value did you gain from implementing JBoss solutions and how did this impact your business? (e.g. improved ROI, increased competitive advantage, better time to market, etc.)

• JBoss Application Server provided us with lowered costs, mainly through third-party licensing costs.
• We’ve saved a lot of development time with the combination of JBoss Hibernate, Red Hat Enterprise Linux, and Oracle because the combination provided us with the ability to quickly understand and resolve any issues that arose during our platform migration.
• 24×7x365 availability is less costly to maintain because of the transparency inherent in JBoss’ open source technology. With transparency, we are able to deal with any issues in and take immediate corrective action to minimize the impact to our clients.
• We’ve experienced the benefits of the speed and efficiency of the Red Hat and JBoss support teams, in comparison to other vendors.

Please provide a technical description of implementation, including the size of deployment. (I.e. Hardware specs, applications, O/S, databases, etc.)

Our IT architecture solution includes 30 HP Blade Servers utilizing Sun JVM 1.4.2 and JBoss Application Server 3.2 on Windows. We also use Oracle databases running in a cluster on Red Hat Enterprise Linux 4.1 on Dell hardware with EMC. For our migration, our CTO played the role of architect and guided the technology.

For the future, we plan to additionally migrate to JBoss ESB and we’re currently in the early stages of exploring JBoss Rules. We also hope to incorporate a monitoring tool like JBoss Operation Network.

Did you leverage JBoss support services, training, or consulting? If so, please describe your experience?

We used migration consulting when deploying JBoss solutions and had a very positive experience. While our CTO acted as the architect, we utilized JBoss consulting for a few months during the migration. Now, we use 24×7 enterprise-level support and have found that any questions we may have are easily answered by knowledgeable support staff. Our experience with Red Hat and JBoss support has been excellent. After several months in production, we ran into an obscure cluster problem and were able to immediately get the assistance of the senior developer from JBoss to help us understand the issues. That level of expertise truly made a difference for our system. We also ran into several Oracle issues while porting our application. Once the problem was identified, Oracle turned around a patch for Red Hat Enterprise Linux in four days. Our development team currently develops on Windows desktops and the same patch that we had for Red Hat Enterprise Linux would only be available for Windows in the next release of Oracle.

Do you have advice for other companies facing a similar business challenge?

Our advice to companies in our industry with the same challenge of needing a reliable, scalable, and affordable solution is to explore open source, specifically Red Hat and JBoss solutions, because this technology is 24×7 stable, provides a full feature set, can be developed on quickly, and maintained with less cost. We’ve saved hundreds of thousands of dollars in fees and even more in development time.

Category: Joint JBoss/Red Hat Deployment
Winner: CompuCredit Corporation
Submitted by: Guido F. Sacchi, CIO and SVP, Corporate Strategies, Cindy Hayden, Manager of Real-Time Integration
Industry: Financial services
Geography: Atlanta, GA

Overview
Used JBoss Enterprise Application Platform to construct an XML Gateway that serves as the company’s main real-time transaction hub—enabling rapid growth, increased productivity, faster service, and millions of dollars in cost-savings.

Please describe your company. (Number of employees, private/public, industry, etc.)

CompuCredit Corporation is a leading provider of credit and related financial services and products for the underserved consumer credit market. Offering branded credit cards and other fee-based products, the company’s strategic competitive advantage lies in its ability to serve creditworthy consumers in a market segment often bypassed by traditional financial institution. For the year ended December 31, 2006, the company had over 4.3 million customer accounts with an aggregate managed portfolio of $2.81 billion in receivables.

Please describe the business and/or technical challenges you faced in this project.

In 2006, CompuCredit was facing two challenges: one business related and the other technical. From the business point of view, our traditional method of using a direct mail campaign of pre-approved credit cards to attract customers had dried up, and we needed to exploit innovative new sales channels such as Internet-based marketing and real-time telemarketing. The ITA (Invitation to Apply) sales channel had a much more significant potential for growth, but required a platform where information was instantaneously available in real-time in order to effectively turn qualified leads into real customers.

Our second challenge was the real-time interoperability of many fragmented applications that existed in our organization. The applications ranged from “Fat Client” type to web-browser based, and were written in a variety of languages such as C, Java, Python, VB, and .net. Our ultimate goal was to create a seamless experience for users and customers (at the presentation or application layer), despite the fragmentation of the application and data sources.

What was the desired solution?

We wanted to create a solution that would quickly obtain data from different areas of CompuCredit, as well as systems outside of the company, and aggregate that data into a single view for the many different business units. It needed to provide interfaces for a variety of applications, such as Real-Time Telemarketing (RTTM), credit card acquisition, on-line services, and even IVR (Interactive Voice Response). The solution also needed to have scalable, reusable components that could be used by all of the different business units at CompuCredit.

Please describe your vendor selection process and why you chose JBoss in the end.

In the past, CompuCredit had been primarily a Solaris™ shop. We had been running Apache but had to eliminate it since it only provided pure web services. When we started with the project, we looked at the available solutions for Electronic Services Busses (ESB), but couldn’t find anything in the marketplace that met our needs at the time.

We originally became aware of the JBoss solutions by reading press reports. We had previous experience with BAE’s WebLogic™ and IBM’s WebSphere™, but getting JBoss up and running was much simpler. These proprietary (non-open source) programs took a lot of configuration to get them working the first time. JBoss worked right out of the box, although we have done some configuration over time to streamline and optimize the system. Plus, we liked the concept of having a “one stop shop” for support of both Red Hat and JBoss solutions.

The last, and possibly most important factor, was that JBoss was open source. We like open source products for their ease of use and the availability of the open-source developer community. While we have not yet had the need to modify the source code, knowing that we can if we need to tweak something is a big plus. This is especially important for innovative projects, such as the new application and frameworks that we are building.

Describe the application you built using JBoss. What role did JBoss and/or JBoss products play in the final solution?

The XML Gateway is a platform that uses all open-source applications with JBoss Enterprise Application Platform as the core. We were able to create a system that fulfilled our vision for the seamless provisioning of real-time information. The system now masks the complexity and fragmentation of data sources and applications. The XML Gateway was developed in conjunction with a network of partners that each brought their individual expertise to the project, but CompuCredit was the driver in the project.

The CompuCredit web framework relies on the XML gateway as one of the main transaction processing engines for all of our real-time services. The solution we created is similar to a home-made ESB that is focused particularly on real-time and web services. We believe that this is a very innovative approach to solving our problems. In the future, we want to look into leveraging JBoss solutions, not just for the XML Gateway, but for a wide range of applications within the company.

What value did you gain from implementing JBoss solutions and how did this impact your business? (e.g. improved ROI, increased competitive advantage, better time to market, etc.)

• The XML Gateway was a key factor in our continued growth. In absolute terms, we have doubled the number of accounts we service in just two years.
• We have also seen tremendous business benefits. Just the improvement in productivity in handling collection and customer service calls saved us between $2 and $4 million per year, and we have seen similar benefits across the entire organization.
• Having reusable components that can be used by all of the different business units at CompuCredit saved us considerable development effort. There are services that are used by all of the different applications, including customer service, telemarketing, collections, online services, customer acquisition via Real-Time Telemarketing (RTTM).
• The time that it takes to process a transaction reduced from minutes or seconds to milliseconds, resulting in increased productivity and greater customer satisfaction.
• The project was easy to implement, taking only eight months from beginning to end and involving the efforts of only four people: one full-time developer and three contractors.

Please provide a technical description of implementation, including the size of deployment. (I.e. Hardware specs, applications, O/S, databases, etc.)

The XML Gateway is a Service-Oriented Architecture (SOA) platform that uses all open-source applications, such as JBoss. It consists of three hefty Dell appliances running Red Hat Enterprise Linux Advanced Server, Sun JDK 1.5, and JBoss Enterprise Application Platform We use Oracle as the primary database, but with some SQL Server. Each server runs two real-time incidences of JBoss and two hot-failover instances should they be needed.

Our application is meta-data driven. We built a large-scale User Interface (UI) for testing and configuration that allowed us to exercise the web services before exposing them to the customer. Then we were able to easily modify that UI to meet the needs of the individual users.

Our platform serviced 100,000 transactions on the first day, and is currently serving 4 million transactions per day. The average transaction time is only a few hundred milliseconds, and the performance continues to improve over time as pieces are honed and hardware improved.

Did you leverage JBoss support services, training, or consulting? If so, please describe your experience?

We have had several people on our team participate in JBoss training and plan to send more in the future. We are willing to make a commitment in training our own people so we can become proficient in developing and maintaining systems of this complexity. We feel that it is a good investment to have this capability in-house. We also used some of Red Hat’s consultants to help us “tune” our environment for peak performance.

We have been a support customer since 2005 and currently have a 24×7 Premium support subscription. We are pleased with our support experience. Everytime we have had to use it, the answers are good and quick.

Do you have advice for other companies facing a similar business challenge?

We are firm believers in open source. The common myth about open source is that it is harder to get up and running. We have found that the truth is quite the opposite.

You also need to take risks and work through the initial investment, which pays off in the long run. Writing the first web service is a lot of work, but once that is done, each additional service is easy.

Category: Emerging and Leading Edge Technologies
Winner:Big Lots
Submitted by: Big Lots Application Architecture Team
Industry: Retail
Geography: Columbus, OH

Overview
Big Lots used the JBoss Enterprise Application Platform and JBoss Seam to develop and implement an interactive Inventory Management System usable across a variety of devices (Point-of-Sale, mobile handheld, and back-office server) in 705 retail stores across the US (rollout to its fleet of approximately 1,350 stores will complete in 2008). The Inventory Management System is part of a multi-year store technology refresh project. Thus far, the changes have resulted in increased productivity at the front end, improved customer experience, and improved ability to respond to future business initiatives.

Please describe your company. (Number of employees, private/public, industry, etc.)

Big Lots is a Fortune 500 closeout retailer with over $4.5 billion in annual sales. With seven distribution centers and approximately 1,350 stores averaging nearly 30,000 sq. ft. each. The company is one of the largest specialty retailers in the country. Big Lots has been a publicly traded company on the New York Stock Exchange for over 20 years.

Please describe the business and/or technical challenges you faced in this project.

Big Lots is two years into a three-year project to upgrade the hardware and software infrastructure at each of its approximately 1,350 stores across the United States. The second phase, completed in 2007, consisted of development and deployment of a suite of inventory management applications used daily at each store. Our legacy electronic cash register system was very limiting from a functionality perspective. In addition, our inventory applications ran on a variety of servers, used differing technologies, and had inconsistent user interfaces, making employee training difficult. We needed a modern solution that would improve system functionality and organizational productivity, as well as enhance the customer experience.

Due to the large number of stores, our implementation also needed to be cost effective. Store bandwidth also presented a challenge, since a vast majority of our stores have very limited bandwidth to the central office. Any new solution needed to provide a great deal of store autonomy and consider the limited bandwidth constraints. Finally, we needed our inventory applications to be usable on a variety of platforms with widely varying input capabilities, everything from a traditional server PC to a touch-screen Point-of-Sale system to a small-screen, pen-based mobile scanning device.

What was the desired solution?

We needed a custom application framework solution that would allow for quick development ‘sprints,’ enabling our development team to quickly add or change functionality to the store applications as necessary to further our business needs. The application framework solution also needed to be accessible from the variety of devices in the store. The solution needed to encompass the needs of both the corporate office and the retail stores.

Please describe your vendor selection process and why you chose JBoss in the end.

We recently switched our enterprise application server of choice to JBoss for development of JEE applications. We wanted to have a consistent direction for both corporate and store applications. We evaluated a variety of different platforms and settled on JBoss Enterprise Platform as the application server. Total cost of ownership and functionality were two of the key components of our decision.

The synergy between what we were looking for and the capabilities that Seam provided out of the box were a good fit, especially with its integration with RichFaces and AJAX4JSF (Java Server Faces).

Describe the application you built using JBoss. What role did JBoss and/or JBoss products play in the final solution?

We run a cluster of four JBoss Enterprise Application Platform servers in the central office, and each store has a back-office server running a single instance of JBoss Enterprise Application Platform server. The servers host a variety of business applications including inventory management, centralized returns, operations dashboard etc. Integration between the server at the store and the cluster in the central office is accomplished through both REST calls as well as a custom data synchronization framework.

We use JBoss Seam-based applications to integrate inventory management functions such as Inventory Transfers, Inventory Counts, Return to Vendor, Item Inquiry, Vendor Receiving, Store Supply Ordering, Inventory Adjustments etc. An example of one of the integrated functions that JBoss Seam enables is Item Inquiry. The store cashiers can open up the Item Inquiry application on their Point of Sale system and query an item using the SKU, UPC or Item Description or on their handheld device or the back-office server. JBoss Seam enables the Item Inquiry application to integrate with the Central Office and provide an accurate count of the number of items in the store or a sister store, as well as across the district and region.

Similarly, when we conduct inventory counts, JBoss Seam is designed to enable the Inventory Counts application to be accessed on handheld devices. A store associate may scan the bar code and the system automatically computes the total number of counts per SKU based on pending transfers, incoming shipments, inventory shrinkage, sales etc. This information is then sent to the central office to report the sales per store and the items that have been moving.

JBoss Seam was used to deliver our Inventory Management applications on multiple devices and browsers, including the servers in our back office, Point-of-Sale (POS) systems, and hand-held devices with integrated bar code scanners on the retail floor. Whether our employees work with a touch screen (on the Point of Sale) or the back-office server or the handheld device, we had to make our applications deliver the same functionality regardless of the interface.

What value did you gain from implementing JBoss solutions and how did this impact your business? (e.g. improved ROI, increased competitive advantage, better time to market, etc.)

• When the rollout of our new POS is complete, the employees will have a highly usable suite of inventory applications that replace a variety of applications built on a myriad of technologies. This has resulted in a reduction in training time and improved productivity.
• Each store has better control over its inventory and processes, and we now have the ability to deploy robust, new inventory-related features.
• Because of the faster throughput and reliability of the platform, stores are also able to operate with fewer registers and servers.
• We were able to consolidate the technology footprint in both the central office and the stores by reducing the amount of hardware and software platforms in use.
• Because of the availability of real time inventory information, customers have a friendlier store experience.
• IT is able to spend less time maintaining application infrastructure and more time developing business logic for the applications. We have improved the construction and design process and allowed new functionality to be developed in less time than with our legacy applications.
• By using open source software the company was able to reduce the cost of deployment in the stores and central office.

Please provide a technical description of implementation, including the size of deployment. (I.e. Hardware specs, applications, O/S, databases, etc.)

So far, we have deployed the new JBoss-based system to 705 of our 1,350 stores, and we expect the deployment to be complete by the middle of next year. Each store has an IBM store-hardened server with an Intel 3 GHz CPU and 2 GB of RAM running 64-bit Windows 2003 Server. We also have IBM electronic cash registers that host an Oracle POS system from which we launch a browser for accessing the JBoss Seam-based inventory applications.

Our hand-held devices are Symbol MC3000 Mobile Computers with 64 MB of memory. They communicate with the server over a wireless network. The handheld device allows employees to access most of the inventory application functionality using a built-in bar code scanner and the numeric keypad.

At the central office, we have a cluster of four JBoss Enterprise Application Platform servers running on two physical machines. These are typical Intel-based 64-bit Windows 2003 servers.

Did you leverage JBoss support services, training, or consulting? If so, please describe your experience?

We have had no specific JBoss training because we haven’t felt the need for it. We did use a JBoss consultant as part of our initial contract to help us set up the initial clusters.

We have used both the official support we receive through our JBoss Premium Support subscription, as well as the less formal, but invaluable, support from the open source community.

Do you have advice for other companies facing a similar business challenge?

Invest in a team of quality people; spend time researching new technologies to be assured of using the right tools for the job. Develop a simple open source policy that covers both consumption and contribution. And then contribute by testing, patching, and even developing new software. When developing custom UI components, work with the toolkit providers so that you don’t duplicate efforts.

Copyright © 2007 Red Hat, Inc. All rights reserved.

Category: New Generation Technology
Winner:Lexicon Genetics
Submitted by: Buckley Kohlhauff, Mark Ma, Jason Williams
Industry: Biotechnology
Geography: The Woodlands, Texas

Overview
Selected for their use of JBoss Seam to glue together Hibernate, JSF, EJB3, and JBoss jBPM to dramatically simplify their development process and create a robust platform that can deploy mission-critical applications for the Texas Institute of Genomic Medicine.

Download JBoss Innovation Award Submission
Download JBoss World Las Vegas Presentation

1. Please describe your company. (Number of employees, private/public, industry, etc.)
Lexicon Genetics is focused on the discovery of breakthrough treatments for human disease. We use our proprietary gene knockout technology to systematically discover the physiological and behavioral functions of genes to identify potential drug targets. We have advanced more than 70 knockout-validated targets into drug discovery programs.
Lexicon Genetics employs over 700 people between our two sites in The Woodlands, TX and Princeton, NJ. Lexicon’s revenue for 2005 was $76M

2. Please describe the business and/or technical challenges you faced in this project.
The challenge was to re-engineer a legacy production system that has been running for five years. The business logic for the system was spread among various layers and components. Most of the documentation that existed was outdated. We had less than a year to redesign and implement the core architecture and workflows. In addition the new system needed to address the fundamental problems that were present in the existing system and be flexible enough to support the same goals with different business processes.

We needed to reengineer a legacy application from php/apache to an enterprise platform in order to support our major involvement in the recently established Texas Institute for Genomic Medicine (TIGM). We have implemented other projects on the JBoss platform and have been pleased with the results. JBoss is a powerful and stable application server and we feel that the JBoss Seam framework will revolutionize Java Enterprise development.

3. What was the desired solution?
Our guiding philosophy was to select a group of frameworks that prevented us from writing a lot of non-business code, but at the same time allowed us to make modifications quickly if we needed to. In addition we always want to leverage standards in the industry. We have utilized J2EE for 3 years so our solution needed to stay within that context to leverage our internal knowledge and skills.

We selected JBPM in 2005 as our solution for modeling our business processes. We selected JSF as our UI framework since the application needed to be accessed from a browser. The introduction of seam excited us because we felt that too much time was spent on connecting the backend layers in previous JSF applications we had written. We had some internal solutions that were built upon codegeneration, but they weren’t flexible and couldn’t help us with JBPM integration.

The Mouse production software encapsulates a complicated workflow covering many scientific and business processes involved in the production of genetically-modified knockout mice. It has to be flexible enough to meet the needs of a large user-base comprised of many distinctive roles. It also needs to be scalable and configurable enough to be used by other organizations involved in TIGM that may need customized workflows.

4. Please describe your vendor selection process and why you chose JBoss Solutions in the end.
We already had selected JEMS as our stack for enterprise applications in 2004 after reviewing alternatives from BEA, IBM, and Oracle. That decision was based upon a matrix of feature requirements, cost, support options, references, and published data.

We also evaluated other application servers and frameworks such as Oracle JDeveloper/BPEL, JRun, and Spring. JBoss proved to be the most cost-effective and robust provider. JEMs allowed us to quickly adopt SOA-based development, increasing the reusability of our code. It enabled us to break our company’s scientific and business processes down into granular projects that fulfill specific needs and adapt to changing requirements in our fast-paced software development lifecycle.

5. What role did Red Hat and/or JBoss products play in the final solution?
JEMS is our platform for application development. The trend we see is tighter integration with the JEMS suite, therefore we lean towards selecting tools from within the suite.

6. What was the overall impact of the project on your business? (e.g. improved ROI, increased competitive advantage, better time to market, etc.)
The project is directly tied to recognizing revenue as well as providing a competitive advantage for Lexicon, TIGM, and our partners.

The Seam framework significantly reduced development and deployment time by gluing together Hibernate, JSF, EJB3, and JBPM. It enabled us to focus solely on our complicated scientific and business logic without having to put together the pieces of the enterprise framework ourselves. With the traditional Java Enterprise architecture there are so many tiers that have to be explicitly implemented, configured and glued into place. Seam makes all of that transparent to the developer.

We especially benefited from Seam’s introduction of the conversation context, as well as the integration of JBPM. The conversation context helped us resolve classic technical challenges such as users using our software in multiple windows. The JBPM framework allowed us to clearly define our business and scientific processes, and it provided a simple and efficient way of implementing the workflows, while implicitly maintaining the data integrity.

For user interface development, we have been using JSF for 2 years. Seam’s direct integration of JSF made it the perfect framework to allow us to reuse some of our existing custom JSF components that provide a rich user interface for our users.

7. With the savings gained from implementing JEMS, how did you reallocate your cost savings within your company?
The project is directly tied to recognizing revenue as well as providing a competitive advantage for Lexicon, TIGM, and our partners.

8. Please provide a technical description of implementation, including the size of deployment. (i.e. Hardware specs, applications, O/S, databases, etc.)
We have a clustered front-end and back-end running on Dell 2850 servers. We have 4 separate clustered instances of JBOSS spread on 3 servers. Our backend comprises of 3 servers running Oracle 9i RAC.

9. Did you leverage Red Hat support services, training, or consulting? If so, please describe your experience?
We used JBoss support and training. The level of support and training is on par with what we receive from our other vendors. Early adopters clearly have an advantage to effect the direction of the product and therefore benefit from support.

10. Advice to other companies considering JEMS.
Get support and training early in the process. The learning curve isn’t steep but it can be completely avoided by doing what you would normally do with other software purchases.

Selected by the community as the 2006 JBoss Innovator of the Year
Category: Service Orientated Architecture
Winner:RLPTechnologies
Submitted by: RLPTechnologies Team (see below)
Industry: Bio Engineering
Geography: Farmington Hills, Michigan

Overview
Selected for their use of JBoss AS and Hibernate as the foundation for their SOA-based platform that has revolutionized how data is collected, enhanced and compiled to increase data-file processing performance by 70%, increase scalability by 400%, and enrich the timeliness, accuracy and quality of R.L. Polk’s data for the automotive industry.

RLPTechnologies Team: Manoj Bansal, Celeste Castello, Kiran Dattani, Mike Davis, Darrin Deeter, Louis Devaney, Rick Drape, Cornell Furtuna, Bill Frost, Indira Harracksingh, Kusunam Srinivas, Joe LaFeir, Norm Marks, Sergey Melnichenko, Kris Musial, Hans Mosher, Kunnummal Naheed, Kathy Northcutt, Prabakhar Oiha, Ivan Provalov, Lawrence Rama, Mike Reed, Gary Rosteck, Clara Sagan, Scott Thibodeau, Kathy Northcutt, Vasconi, Geoff Volpe, Lisa Wagner, Pei Zheng

1. Please describe your company. (Number of employees, private/public, industry, etc.)

R. L. Polk & Co. is the premier provider of automotive information and marketing solutions to the automotive world and its related industries—automotive and commercial vehicle manufacturers and dealers, automotive aftermarket companies, insurance companies, finance companies, media companies, advertising agencies, consulting organizations, government agencies and market research firms.
A privately held global firm, Polk is based in Southfield, Michigan with over 1,300 employees located at operations in Australia, Canada, China, France, Germany, Japan, the Netherlands, Spain, the United Kingdom and the United States.

RLPTechnologies, Inc. is a wholly-owned research and development subsidiary dedicated to deploying world class data-driven technology products that support customers’ needs to turn vast amounts of data into business value with accuracy, speed and security.

RLPTechnologies specializes in building industry-leading data solutions that serve as the foundation for focused, in-depth research, analysis and action across multiple industries, enabling the effectiveness of business intelligence tools and applications that “mine” intelligence from the data.
Our company vision is nothing short of revolutionizing the way data is collected, standardized, enhanced and compiled into a Single Source of Truth. Our solution, the Enterprise Information Factory, does more than just build a consolidate view of data. It also:

• Processes data faster
• Improves data accuracy
• Ensures compliance with regulations and reporting needs
• Reduces the costs to process, support and maintain information assets

We’ve taken a unique approach to building the Enterprise Information Factory (EIF), applying the principles of lean and flexible manufacturing, along with IT industry standards including, Service Oriented Architecture (SOA).

2. Please describe the business and/or technical challenges you faced in this project.

In 2004 Polk’s CEO, President and Executive Committee held a series of strategy meetings to discuss how Polk could first maintain and then improve its competitive advantage amid significant industry, regulatory and technology change.

Over the years, Polk has enjoyed a position as the market leader and is the “gold standard” for automotive vehicle and consumer data. This data is used by every automotive brand to make critical decisions about their businesses. Further, many automotive suppliers, dealers, and other automotive-related businesses (finance and insurance, media, research, government agencies) utilize Polk solutions. Polk’s data and applications are used by its customers to help them make decisions about areas such as dealer and network planning, parts and inventory planning, customer segmentation and target marketing, and vehicle verification to name but a few. Having served the automotive market since 1922, Polk provides data that is ‘court-tested’ to defend franchise decisions made by OEMs. Further, Polk’s data is used for recall purposes to ensure that every vehicle owner is notified of recall campaigns.

Never wanting to “rest on its laurels,” Polk has continuously improved its data management methods over the years. Given today’s environment, in which privacy compliance is introducing even tighter restrictions on how data can be used, the time was right to move beyond continuous improvement to develop a innovative approach that would revolutionize Polk’s core foundational data warehouse.

Polk’s executive leadership had a healthy debate centered on two fundamental issues at the earliest stages of the re-FUEL project.

First, to be successful with a project of this significance, size and scope, the IT team members tasked with accomplishing success would need to be focused fully on this project, and not burdened with other daily demands. In other words, we didn’t want to “change the tires on the car while it was moving.”

Second, Polk realized it was not alone in facing the challenges and complexities inherent in large-scale data management, data warehousing, and application development/integration. According to Gartner, in 2004 organizations were faced with managing 30 times more data than in 1999. This trend is not likely to change.

With both issues in mind, Polk’s senior management concluded that the appropriate course of action was to develop a new subsidiary, RLPTechnologies. The charter for this organization was to develop a working software solution for use by the parent that would also be viable for other organizations.

A plan was devised to totally re-engineer the core revenue generation engine that powers the company and to do it in such a way that it:

• maintains and improves the current competitive advantage for the next 10 years, and
• creates a subsidiary company to “spin out” the technology innovations into the market place.

In December of 2004, the Polk Board of Directors approved the re-engineering program and the creation of RLPTechnologies, a wholly owned subsidiary of R.L. Polk & Co.

The re-engineering program was code-named re-FUEL ( Re-engineering Functions with Urgency, Excellence and Leadership). The re-FUEL vision was described in the charter approved by Polk’s Board of Directors as follows:

“The re-FUEL vision is nothing short of revolutionizing the way data is collected, standardized, enhanced and compiled into data warehouses. …..The solution will be designed with the awareness of today’s security threats and data privacy issues…. The solution will be designed to incorporate a high level of quality automation and statistical trending to detect, and potentially predict, data quality issues….This effort should produce a world class data collection, enhancement, and compilation solution; a system that utilizes superior technologies and methods to produce superior results and profitability. It is not an exercise in continuous improvement, but a journey of discovery and innovation”

In essence the re-FUEL team was given the rare opportunity to take a clean sheet approach to designing the new systems, processes and organization. The business vision was established, and referred to as 50/50/100:

- Lower Total Cost of Ownership

- Improve data processing and timeliness and availability

- Protect Polk’s rich heritage as the industry standard, and provide improvements in identifying problems earlier in the process

The re-FUEL team evaluated and eventually embraced a standards based, service oriented architecture (SOA) as the foundation for the new system. As a new IT architectural paradigm, SOA provides significant benefits relative to protecting legacy investments, reducing costs, and providing accelerated time to development. The team also embraced the principles of lean manufacturing – continuous material flows, standardized process, and eliminating waste – which aligned closely with the 50/50/100 goals.

Over the next 14 months, a team led by the new subsidiary (RLPTechnologies) went through an aggressive project schedule to:

The system has been built, and is being deployed in phases. The project is entering the final phase of parallel operation, which will occur from March through June, 2006. Following this phase, the Polk “data factory” will use the new solution exclusively to manage the wealth of data Polk collects.

A conversion of over 2.5 billion data records from the existing Polk data warehouse will be run through the system for consistency.

The program has delivered on both the business vision (competitive advantage & 50/50/100), and the technology vision of a true service oriented architecture (SOA) – enabling Polk to recognize significant benefits, while leveraging the new system to further strengthen its competitive advantage.

3. What was the desired solution?

The solution (The Enterprise Information Factory) was developed by RLPTechnologies as a comprehensive software application that manages how data is collected, standardized and enhanced, and compiled it into a Single Source Of Truth (SSOT) to feed use in analytical and operational applications. The Enterprise Information Factory innovates in two primary areas: business process and technology.
Business Process Innovation
We’ve learned from the principles of lean manufacturing, and applied those lessons to the area of data processing in the development of the Enterprise Information Factory (EIF). Key lean principles of continuous material flow, process automation, standardization, quality controls and continuous improvement are built into the core of the solution.

The solution handles incoming data in much the same way as a factory built on the principles of lean manufacturing handles raw materials. As soon as inbound data (the EIF’s raw material) arrives, the factory immediately recognizes the availability of data and begins processing it. This automated collection and real-time processing of Polk’s data reduces the overall time for the data to reach its key business intelligence and other transactional business systems. The Enterprise Information Factory has eliminated manual processes, allowing Polk to recognize overall improvements of up to 70 percent on processing inbound data.

This type of real-time automated processing is possible because Polk’s business analysts have the ability to setup, or “tool” the Enterprise Information Factory with custom business rules for the processing of any particular source or type of data. Once configured, the EIF runs as a fully automated system, requiring minimal manual activity. Like any highly automated system, the solution needed a robust monitoring and process control system. An operations management dashboard was built to provide visibility into the health of the factory. The operations management portal displays real-time metrics of the EIF performance. It also provides access to exception queues, allowing analysts to resolve issues that occur during automated processing. When exceptions are encountered, the specific data in question is “pulled off the line” and alerts are sent out. This allows for the continuous flow of all other data moving through the EIF, and makes Polk business analysts immediately aware of the issue, so they can begin prompt resolution. Once an exception is resolved, the data is placed back into the Enterprise Information Factory’s workflow to complete processing.

The EIF may raise exceptions for a number of reasons, but a sophisticated data quality module is the primary source. The data quality module evaluates data content at various check points in the factory, from the time data arrives through to the delivery of finished data products. The data quality tool allows Polk analysts to establish rules on such factors as data consistency, completeness and value distribution. Polk analysts can use business rules to adjust thresholds up and down for acceptable variances in the data. As a result, Polk quickly identifies data quality issues and responds to them early in the data processing lifecycle.

Through the innovative use of lean manufacturing principles in the field of data processing, the Enterprise Information Factory has allowed Polk to recognize significant business process efficiencies in a once very manual process.

Technology Innovation
The re-FUEL program was structured to allow the team to architect the Enterprise Information Factory from a clean sheet of paper, without concern for the technology constraints of existing platforms. This allowed the team to develop a very innovative technology solution. The three key areas of innovation include: the creative use of an enterprise service bus (ESB) as the solution backbone, a custom Service Orchestration engine that provides dynamic integration to web services, and the implementation of a GRID computing platform.

Enterprise Service Bus (ESB)
The ESB serves as the JMS messaging backbone of the Enterprise Information Factory. The ESB is essentially the underlying foundation that holds together all modules of the factory, such as:

• Data Capture
• Standardization
• Data Enhancement
• Quality profiling
• Assembly

And common foundation services, such as:

• Logging
• Exception handling
• Scheduling
• Security

A standard message structure facilitates communication between components in the EIF. This approach provided a tremendous amount of flexibility when developing and integrating components of the solution to create a large composite application.

Service Orchestration Engine
The Enterprise Information Factory is founded on a service oriented architecture. At the center of that SOA is a custom developed service orchestration engine. This engine manages all business services executed against the data moving through the factory. The Service Orchestration engine was specifically designed to handle high volume and a high degree of flexibility for the handling of changes to data and business services.

The solution provides an application that allows Polk business analysts to create and modify service orchestration profiles. These profiles are based on registered services and the type of data that is being processed.

The EIF service orchestration function is based on a data driven dependencies engine (D3E). At run time the service orchestration engine retrieves a profile that defines what services have been assigned to the data source. Through the use of Web Service Description Language (WSDL) in the service registry and the inbound data schema, the engine automatically derives an optimized execution path. Unique parsing, segmenting and aggregation routines developed by RLPTechnologies allow the engine to perform parallel processing and manage calls to and from all services. All communication with registered business services use common web service standards and protocols such as SOAP, JMS, HTTPS, XML and WSDL.

The capability provided by service orchestration allows Polk to quickly integrate business services provided through either the use of commercial off-the-shelf software (COTS), legacy applications, or external providers. Polk integrated business services and used COTS such as DataFlux for name and address cleansing, and iLog for sophisticated VIN rules processing. Additionally, Polk’s legacy business logic in COBOL was wrapped with a web service interface and connected to the factory.

4. Please describe your vendor selection process and why you chose JBoss Solutions in the end.

The selection of JEMS was decided very early in the project. During the early stages of the reFUEL project, the team completed a conceptual and logical architecture. Based on this target architecture several key foundation software and hardware components were identified, which included application server and object/relational persistence amongst others.

The aggressive time frame did not allow the team to do a broad sweep of available products, so the team quickly developed a short list based on market research firms such as Gartner. Based on this research the team evaluated the first round of candidate technology in a lab, to assess stability, maintainability, performance and interoperability. The results of this testing, coupled with the desire for RLPT to eventually develop a commercial product – JBoss was the clear winner.

5. What role did Red Hat and/or JBoss products play in the final solution?

The JBoss Application Server and Hibernate Object/Relational Persistence products were critical components of the technical foundation for the solution.

The JBoss application server is used to run all Java components of the Enterprise Information Factory (EIF) developed using Hibernate. The core EIF Java applications used to configure the operations of the data factory include: Data Capture, Reference Data Management, Data Quality, Service Orchestration Gatekeeper and Assembly functions. These are critical business applications used by Polk Analysts to perform their day to day jobs.

The performance and scalability of the factory was paramount for this project, and was achieved using the JBoss application server. The solution needed to scale to support over 100 transactions per second while processing though several business services. The EIF Service Orchestration Engine (developed in Java), is the foundation of the SOA architecture and manages all data movement, and calls to and from all registered services. In addition to running custom developed components, JBoss application server was also used to run several of the COTS products used in the solution. This included an implementation of iLog jRules business rules execution engine running in JBoss. The business rules implementation for Polk Vehicle decoding contains well over 600,000 rules in multiple rule sets deployed across multiple application servers, one of the largest implementation of rules for a solution using iLog.

6. What was the overall impact of the project on your business? (e.g. improved ROI, increased competitive advantage, better time to market, etc.)
The Polk Executive Committee approved the re-FUEL project as the #1 priority for Polk’s FY05 and FY06 business plan.

At its basis, the re-FUEL project focused on re-engineering and boosting the performance of Polk’s core revenue-generating engine, the power driving the company’s business success. Polk has realized significant business results from the re-FUEL project, including both revenue protection and generation, combined with equally significant cost-savings.

Revenue Generation
The project positively impacted Polk’s revenue picture, both in terms of protecting current revenue streams as well as supporting additional revenue growth.

Revenue Protection
Over 50 percent of Polk’s market-leading automotive data and analytical solutions are supplied by the new solution. Enhancements in the speed, accuracy, and quality of the data, combined with improved regulatory compliance capabilities, have enabled Polk to maintain a position of strength compared to its competitors. Two elements of the 50/50/100 plan—50 Percent Faster and 100 Percent Quality–are worth noting as drivers of significant business benefits for Polk.

50 Percent Faster – Tests to date show improvements of up to 70 percent in data-file processing speed (on average). For example, an average state registration file that previously would have required manual processing by as many as three full-time employees (FTE’s) and four hours of processing time, now is processed in an automated fashion in roughly 23 minutes. Further, RLPT’s approach to grid computing has allowed the solution to scale to process ~100 transactions per second, nearly four times Polk’s average of 25 transactions per second – providing headroom to accommodate, processing spikes, future transactions or business growth.

100 Percent Quality – The standardization and enhancement functionality of the Enterprise Information Factory measurably improves the accuracy and completeness of the data, preventing quality problems that might impact customer satisfaction. Automated data quality checkpoints allow for earlier recognition of problems and enable the team to resolve issues before the data is delivered to Polk’s business intelligence and operational applications. This functionality drives a focus on preventing issues–or at worst, recognizing them early–following the rule of thumb that “It costs $1 to prevent a problem, $10 to identify a problem, and $100 or more to fix it.”

Faster delivery of higher-quality information should translate into improved customer satisfaction, resulting in continued long-term business commitments. Given Polk’s subscription-based models, this will enable continued positive financial returns for the 135 year-old company. Fending off any threats to the core business will allow Polk to maximize new revenue generating opportunities and drive double-digit growth.

Revenue Generation
A significant benefit delivered by the solution for Polk is their ability to shift focus from data management to product strategy and application development. Armed with the flexible environment provided by the Enterprise Information Factory, Polk’s Product Strategy group can look for new data sources to enhance its offerings, while also developing new analytical and operational applications to leverage more timely and complete data. Polk expects these new capabilities to prime the company for future growth, and embolden managers with the knowledge that they will experience reduced time to market in future development efforts. The EIF solution will be rolled out in phases, and the team is currently working on how best to deploy it globally to further strengthen the consistency and completeness of Polk’s data and product applications worldwide.

The formation of RLPTechnologies was founded on the knowledge that the Enterprise Information Factory could also solve challenges facing other large organizations–and generate new revenue streams in the process. Market trends support this approach. A November 2005 Gartner study on data integration, for example, shows that in North America, 21% of respondent companies expect constant investment in data integration, with 60% expecting to increase such investments in 2006. Further, the same study asked respondents about the “degree to which their SOA initiatives included the service orientation of data assets and creation of data services.” Gartner concludes that “with only 37 percent indicating a strong focus on this topic, most organizations appear to be at risk for failure in their SOA efforts because they are probably not addressing fundamental issues, such as consistent transformation, delivery, and quality improvement of the data.”

With its comprehensive approach to data integration and management and with a service oriented architecture at its core, RLPTechnologies solution (the Enterprise Information Factory) is well positioned to capture continued investments by businesses in these areas.

Supported by strong partnerships with industry-leading consulting firms and software technology companies, RLPTechnologies will provide significant growth for the parent company. The three-year business projections are expected to deliver between 5 and 10 percent top-line revenue growth for Polk.

Cost-Savings
The first element of the business vision that was established in the 50/50/100 plan was 50 percent more efficient. The re-FUEL project allowed this goal to be met, with significant cost benefits to be realized by Polk in two core areas:

Leaner, Better Aligned Team
Prior to the re-FUEL project, Polk’s previously-named Data Operations team had moved from Cincinnati to Polk Headquarters in Detroit in 2003, resulting in a centralization of IT functions. The re-FUEL project transformed the structure and size of this group, creating a more cost efficient and focused unit. Renamed the Data Factory, the group is now 43 percent smaller, and team members have significantly different roles and responsibilities. The group is structured more efficiently, with roles that align directly to the functions of the solution (Data Capture, Standardization & Enhancement, Reference Data Management, Single Source of Truth/Operational Data Store, Assembly, and Operations Management). Further, the reduction in manual processes has enabled the group to focus on strategic management and analysis of data, including areas such as issue resolution and handling.

Lower IT Operating Costs
The implementation of RLPT’s grid computing model will result in significant savings for Polk. By moving away from a mainframe-based system, the grid will operate with hardware costs that are 65 percent less. This change amounts to savings of millions of dollars per year for Polk. Finally, improvements made to the open systems environment are leading to additional savings of 30 percent per year compared to prior operating budgets.

7. With the savings gained from implementing JEMS, how did you reallocate your cost savings within your company?
The savings gained from the project including those from the implementation of JEMS are invested back into the business to drive product development efforts to strengthen Polk’s competitive advantage as the market leader in business intelligence for the automotive industry.

8. Please provide a technical description of implementation, including the size of deployment. (i.e. Hardware specs, applications, O/S, databases, etc.)
GRID Computing Platform
The technology stack for the Enterprise Information Factory operates in a grid computing environment running Linux Redhat on Intel Xeon processors. The target production grid is comprised of 49 servers and 118 processors. The database holding the Single Source of Truth contained in a 4.5TB database with over 2.5 Billion transactions. The operation and management of the grid is accomplished through the combination of JBoss clustering, the EIF Service Orchestration and Oracle 10g GRID.

Embedded in the EIF solution is leading commercial off-the-shelf software (COTS); Oracle 10g database grid, portal and Oblix, Tibco BusinessWorks, Dataflux dfPower Studio and iLog jRules to accelerate the time to market for the solution. The EIF application that wraps all of these technologies together is a series of J2EE applications running in clustered JBoss application servers.

The grid based computing platform has allowed both significant cost savings and flexible scalability options to provide capacity on demand.

9. Did you leverage Red Hat support services, training, or consulting? If so, please describe your experience?

Yes, we engaged JBoss for support, training and consulting. The development team ramped up very fast and a number of outside contractors were used. To ensure proper use of Hibernate we purchased training programs for our team. Additionally, we used JBoss consulting to provide assistance with tuning activities.

10. Do you have advice for other companies facing a similar business challenge?

Establish a sound SOA architecture up front and stay true to it as much as practical.

Category: Clustering
Winner: J. Craig Venter Institute
Submitted by: Pete Davies, Indresh Singh, Tom Dolafi, Chris Lemieux, Sean Murphy, Adam Resnick, Angelo Trivelli, Bryan Yu, and Saul Kravitz
Industry: Bio Engineering
Geography: Rockville, MD

Overview
Selected for use of JBoss messaging and clustering to provide the stability and scalability necessary to process in excess of 40 million traces in batch across a 2 node cluster that supports over 100 DNA sequencers (scaling to 8 nodes to process large collections of traces) while also saving the not-for-profit genomic research center over $500,000 per year in licensing and maintenance costs.

Download JBoss Innovation Award Submission

1. Please describe your company. (Number of employees, private/public, industry, etc.)

J. Craig Venter Institute is a not-for-profit research institute dedicated to the advancement of the science of genomics; the understanding of its implications for society; and the communication of those results to the scientific community, the public, and policymakers. Founded by J. Craig Venter, Ph.D., the Institute is home to approximately 200 staff and scientists with expertise in human and evolutionary biology, genetics, genomic and environmental policy research.

2. Please describe the business and/or technical challenges you faced in this project.

The J. Craig Venter Institute’s Joint Technology Center (JTC) opened in June 2003. The facility was designed to be one of the world’s leading DNA sequencing organizations, providing DNA sequencing and resequencing services for the Venter Institute and collaborators worldwide. The JTC executes 150-200 projects a year, 45-50 concurrently, and has a capacity of 80 million sequence reads (lanes) per year. The facility was designed to scale to 320 million lanes per year. Producing high quality data from a 24×7 high throughput DNA sequencing facility servicing many concurrent projects requires a comprehensive and robust Laboratory Information Management System (LIMS).

The challenge was to build a LIMS supporting the JTC’s evolving lab processes that would provide the performance, scalability, and robustness the JTC’s high capacity operation, with materials tracking, workflow, and process control. In addition, the LIMS was required to support integration with hundreds of laboratory instruments, computational pipelines, and analysis tools; real time quality control reports and data delivery to collaborators and public data repositories. We evaluated COTS LIMS tools, but all fell well short of the JTC’s LIMS requirements. The primary shortcomings were performance at the transaction and data volumes contemplated, and the ability to integrate with analysis pipelines and instrumentation.

3. What was the desired solution?

To track all the lab processes and the data, we designed and implemented a suite of LIMS applications, providing bar-coded tracking of users, reagents, and material transfers throughout the facility. The LIMS is built on an Oracle 10g database, with a rich stored procedure layer. Business logic is implemented in EJBs, and most user interaction is via JSPs. The first version of the LIMS was deployed in April 2004.

Extensive integration of laboratory robotics has been facilitated by the use of J2EE technology provided by JBoss. Our ~110 DNA sequencers each host a JBoss instance, supporting user interactions via JSPs, data transfer and remote monitoring via HTTP, and interaction with the LIMS via JMS and EJBs. We have integrated fluid handling robots used to setup experiments in a similar manner. Symbol wireless PDAs have been integrated to track steps where lab staff mobility was critical.
DNA sequence data flows from the sequencers to a pipeline where it is reduced and analyzed, and loaded into the LIMS database. The pipeline is built using both JMS and EJB technology and integrated with both our in-house Oracle database, as well as a Sybase database from one of our key customers. The robustness of this pipeline in the face of DB connection stability enabled by the J2EE features as embodied in

JBoss has produced dramatic increases in customer satisfaction as well as greatly reduced production IT overhead.

Suites of LIMS applications are deployed on three, 2 node JBoss clusters. For on demand, batch processing, we have an 8 node JBoss cluster, which can easily process 40 million traces within days.

We are leveraging JBossMQ with HAJMS to process traces with a JBoss cluster using a fork and join approach. We used this approach to parallelize trace processing for high performance. When the TraceProcessing MDB receives a trace processing request, it forks the job into many smaller jobs by sending messages into another JMS queue. These messages are received by a second MDB and the smaller jobs are then executed on all clustered servers in parallel. The TraceProcessing MDB waits for completion messages from all of the MDBs which are processing the smaller jobs. After receiving the completion messages from all of the jobs, the TraceProcessing MDB aggregates all results and sends a message into another queue which is processed by a Loading MDB to persist the data. This is a very simple, stable and highly scalable approach to process millions of traces. In case we need to add more computing power to our JTrace server farm, we simply add new JBoss nodes to existing JBoss cluster.

4. Please describe your vendor selection process and why you chose JBoss Solutions in the end.

Members of our team had experience using JBoss since 2000. Once we made our strategic decision to heavily invest in J2EE, JBoss was our first choice. The main considerations were its feature set, quality, cost, and clear development roadmap.

5. What role did Red Hat and/or JBoss products play in the final solution?

Building on JBoss has dramatically reduced the development time for our LIMS, and increased its stability. The ability to deploy Java application servers to all of our DNA sequencers has greatly improved the cost effectiveness of our LIMS integration with the sequencers. JBoss JMS and Clustering provided the vertical scalability to our application; we setup an 8 node JBoss cluster to process large bundles (40 million) of traces in batch.

6. What was the overall impact of the project on your business? (e.g. improved ROI, increased competitive advantage, better time to market, etc.)

We currently deploy about 140 JBoss application servers in production, using advanced features like HAJMS, clustering, cache etc. We have probably saved more then $500K per year on application server licensing costs by using JBoss. JBoss free, open source applications provide the unmatched scalability, where we can add new JBoss instances without worrying about any licensing cost.

7. With the savings gained from implementing JEMS, how did you reallocate your cost savings within your company?
Any software dollars saved through the use of JBoss have been reallocated to development software that is core to our scientific mission.

8. Please provide a technical description of implementation, including the size of deployment. (i.e. Hardware specs, applications, O/S, databases, etc.)
DNA Sequencers: 100 ABI 3730xl and 6 ABI 3100 DNA Sequencers
Integration with ABI 9700 Thermalcycler
Beckman Biomek® FX Laboratory Automation Workstation
Hudson Controls Laboratory Automation Stackers and Print and Apply Stations
Zebra Barcode Printers and Scanners
Application servers run under Linux on HP BL20p blade servers
DB Server runs on a two node Linux cluster (HP DL580s).
Symbol PPT-8846 Wireless Pocket-PC based PDAs with scanner
3 Oracle 10gRAC cluster with more then 1Terrabyte of data.
JBoss 3.2.6 Java Application Servers
Panscopic Scope Server Reporting Tool ((www.panscopic.com)
Bugzero Issue Tracking System (www.websina.com)
DiskXTender from Legato
Celera Assembler
Custom Bioinformatics Software developed at the Venter Institute

9. Did you leverage Red Hat support services, training, or consulting? If so, please describe your experience?

Yes, we used JBoss advance training provided JBoss and we found it very good. Training added a lot of value to our employees.

10. Do you have advice for other companies facing a similar business challenge?

JBoss provided excellent enterprise applications which are open, stable, and scalable. We scaled from 1 standalone server to an 8 node clustered sever with some trivial changes. JBoss stability and scalability is exceptional. Why pay for other black box applications server when you can get JBoss open source applications, which can add value while drastically cutting costs.

Category: Core Infrastructure

Winner: Cendant Distribution Travel Services Group, Inc

Submitted by: Bryan Harwood & Chuck Clark

Industry: Travel

Geography: Chicago, IL

Overview
Selected for their use of JBoss AS and its JMX capabilities as the foundation for their services container that allows them to provision core travel services more quickly and efficiently to a number of leading travel sites including Orbitz.com and Cheaptickets.com.

---

Download JBoss Innovation Award Submission

1. Please describe your company. (Number of employees, private/public, industry, etc.)

Orbitz was founded in 2000 by five leading airlines. The orbitz.com site launched in 2001 and has become one of the top 3 travel sites. Orbitz was purchased by Cendant Corporation in November 2004.

Cendant Distribution Travel Services Group, Inc. (“TDS”), the entity that oversees Orbitz and the other travel distribution companies at Cendant, is the official support customer for JBoss. TDS made the strategic decision to roll JBoss across all of its platforms, which supports the following websites:

• Orbitz.com
• Cheaptickets.com
• Travelport.com
• Lodging.com
• Orbitzforbusiness.com
• Parts of American Airlines’ website
• Parts of United Airlines’ website
• Parts of Northwest Airlines’ website
• A number of other third-party ‘white label’ sites

2. Please describe the business and/or technical challenges you faced in this project.

Over the past six years, TDS has developed hundreds of custom, scalable, Jini applications that ran within a home grown service container that allowed them to monitor and provision new instances as necessary. These services spanned a number of travel booking needs including rate, re-price, and availability requests as well as the actual booking transactions themselves. The collective services ran across a large farm of several hundred servers.

When Orbitz became part of the TDS business in 2004, TDS determined that it would need to expand the number of travel websites in deployment. The existing home grown services container could not handle the additional complexity so TDS began searching for a new services platform for hosting the growing number of applications. For external-facing services, the TDS team also identified a need to expose these applications as Web services.

From a technical perspective, TDS was primarily interested in a modular container that could expose services with Jini and managed via JMX and that could deploy their SARs easily. This allows TDS to minimize the complexity and manpower needed to run our Network Operations Center. This was a very important point for TDS – the need to run dozens of websites consisting of hundreds of travel services, across a massive server farm, which could be provisioned quickly and easily as needed, without requiring much, if any, manual intervention from the Operations staff.

3. What was the desired solution?

TDS chose JBoss Application Server for its JMX capabilities first and foremost. Because most of the applications are Jini-based, there wasn’t a strong need for a Servlet / EJB / JSP container. Instead, we were primarily interested in a lightweight JMX-based container. The innovative capability for the JBoss AS container to deploy SARs was also a major selling point and will have a major impact on the future management costs associated with the services.

Unlike before when the Orbitz home grown container was in use, TDS can now rely on JBoss for any and all bug fixes associated with the container. Previously. our developers were primarily focused on creating new business-critical services and the core platform container was often neglected. As a JBoss Subscription customer, TDS can now focus all of its attention on building new services and rely on JBoss to continue to innovate and support the core container. This has a positive impact on the business.

The open source nature of JBoss AS also means that our developers get direct access to the code. This is extremely important. It means that a TDS developer can read the source code, understand the inner workings of the container, and correct any application flaws in much less time compared to using a closed source platform. It’s not unusual for a TDS developer to get a call at 3AM if a service is down and the ability for this developer to have access to both the source code and to the technical experts at JBoss has proven to be invaluable.

4. Please describe your vendor selection process and why you chose JBoss Solutions in the end.

Open source software has played an important role in the ability of Orbitz to compete successfully with entrenched companies that are much larger and much better funded. When reviewing application servers, an open source solution was a priority. When comparing the contenders, Orbitz liked the following attributes of the JBoss Application Sever resulting in their JBoss selection:

Technology – Wanted a JMX-based container, ability to deploy SARs easily, and needed the container to run on JDK 1.5, which was uncommon at this particular time.

Support – As one of the leading travel sites of they world, Orbitz understands extremely well the cost of service downtime. As such, developers will be paged if a flaw is discovered – whether that be during normal business operations or at 3AM. It is important to Orbitz that these developers have access to the source code, a large web-based knowledgebase, and direct access to the real experts behind the code.

Financial – In a notoriously competitive market with thin margins already, cost savings was a major factor when selecting an application server. Considering that the Orbitz Web and Middle-tier consists of hundreds of dual-processor servers running multiple JVMs and instances of their services means that open source has saved Orbitz an extremely large amount of money in software licenses.

Open Source Community Involvement – with a unique deployment they wanted to ensure the roadmap fit with their direction. They wanted to us an open source product where they had the opportunity to contribute and influence the future of a project by interacting with project owners and ability to submit feedback.

5. What role did Red Hat and/or JBoss products play in the final solution?

See section 3 for more details on the role of JBoss Application Server. Hibernate is also used for persistence.

6. What was the overall impact of the project on your business? (e.g. improved ROI, increased competitive advantage, better time to market, etc.)

TDS has noticed ROI in operational efficiencies. JBoss AS has allowed TDS to provision services very quickly and efficiently and allows TDS to manage a large Web and Middleware farm consisting of several hundred servers running several hundred services with only a very small Network Operations Center staff. As the TDS business has grown and technical complexity has increased, TDS has not had to increase the operations management staff and costs accordingly.

Thereare also major cost savings associated with eliminating all Middleware software licenses for this large server farm. TDS does not make public specific financial savings however.

Additionally, the open source nature of JBoss AS means that TDS developers have the chance to read the source code and understand the true reasons for any application performance issues. This has allowed TDS to dramatically reduce downtime compared to using a closed source, commercial product.

7. With the savings gained from implementing JEMS, how did you reallocate your cost savings within your company?
The cost savings that came from eliminating software licenses have made an impact to TDS’s bottom line. Additionally, by eliminating the need to further build out and maintain the core platform, TDS can now focus more of its application development team on creating new business-relevant services.

8. Please provide a technical description of implementation, including the size of deployment. (i.e. Hardware specs, applications, O/S, databases, etc.)
Servers: Rackable hardware, mostly X86 with dual AMD processors each running multiple JVMs
OS: Linux (Red Hat 3.0) throughout
Database: Primarily Oracle but also some MySQL and Postgres
Application Server: JBoss AS
See the JBoss Innovation Award Submission for a diagram

9. Did you leverage Red Hat support services, training, or consulting? If so, please describe your experience?

TDS is an Enterprise Support Customer. TDS’s experience with JBoss support has been positive thus far. Early interaction was important with JBoss Sales Manager and consultants because of their atypical Jini deployment.

10. Do you have advice for other companies facing a similar business challenge?

JBoss has proven to be a solid technology solution at TDS and we believe open source solutions such as JBoss have contributed to our success in the technology sector. Companies need to realize the tremendous advantage open source can contribute, such as reduced licensing cost and access to source code. Our advice is to not be tentative and “embrace open source.”

]]>