IT Briefcase Exclusive Interview: Unified, Complete, and Open Data Integration with TalendNovember 6, 2013 No Comments
In the below interview, Yves de Montcheuil from Talend outlines key integration challenges that are arising with the evolution of big data and cloud computing, and the ways in which a well rounded open source data integration solution can help overcome some of these challenges.
- Q. In your opinion, is open source still as relevant to the Enterprise as it was 10 years ago? And if so, why?
A. Open source is more relevant than ever to the enterprise but in a sense, that makes it a lot less relevant. It is more relevant because it’s pervasive and ubiquitous. All enterprise IT stacks make extensive use of open source, and even what were traditional strongholds of proprietary stacks are now penetrated with open source. Innovation occurs primarily as open source, as big data technologies such as Hadoop, NoSQL, and cloud technologies like OpenStack, prove on a daily basis. But this ubiquity also makes it less relevant. It does not matter today if a technology is open source or not. Technologies are selected on their merits, of which the flexibility, the community, the extensibility are part. But nobody chooses open source because it’s open source!
- Q. Application integration is often viewed as a separate discipline than data integration, but should it be?
A. Application integration and data integration serve the same purpose: deliver timely, accurate and consistent data to the business. The distinction between these is a technical one: application integration often works at the application API level, invoking application methods, while data integration connects directly to the data repositories. Both techniques have their benefits and drawbacks, but they also address different use cases in terms of latency, volume, transactional integrity, etc. Because of the multiplicity of these use cases, these techniques are better leveraged when they can be used together.
- Q. What are the biggest data integration challenges your clients are bringing to the table, and how is Talend working to overcome these challenges?
A. We see two major challenges our customers need to address. The first challenge is the “ubiquitous data,”making all data available to the business – people, applications, processes, etc. – no matter where this data resides, how dynamic it is, and how complex it is. Getting to this data, reconciling it, and making it available in the form that consumers need (persistent or not, real-time or not, via services, batch, etc.) is a key challenge for any type and size of organization. The second challenge is a consequence of the first one, and it is a result of the demand to leverage “new data”. Some call it “big data” but the term is not entirely accurate in this context, since it includes dark data, social data, public/open data, the Internet of Things, etc. This new data adds extreme complexity and scalability issues to the primary “ubiquitous data” challenge.
- Q. How do big data and cloud computing create new integration challenges for enterprises today?
A. It’s not so much big data, but rather new data that creates unique challenges. Many organizations have been dealing with big data for some time, and the new technology platforms available today enable the addition of the new data, essentially turning “big data” into “total data”. The access methods, the form of this data, its volume, all add to the complexity of the equation.
Cloud computing, not widely adopted, creates different types of challenges in terms of distribution, latency, bandwidth, access methods. Applications and systems are not on-premises or in the cloud, they are both on-premises and in the cloud. This complex hybrid environment is the new everyday reality of integration teams.
- Q. As BYOD continues to gain a presence in the workplace, do you feel that design, development and integration of mobile extensions to business applications are crucial to gaining the business value promised by mobility?
A. Mobile applications are not only critical, but they also add new complexities to integration. Data needs to be available to anyone at any time, and this means also on web portals, smartphones and tablets. Users don’t want to compromise performance or breadth of functionality because they use a different system. But application and integration designers need to adapt to different form factors, storage capacity, bandwidth, data plans, and other constraints that force them to rethink data access for these platforms. At the end of the day, truly efficient integration architecture addresses these challenges.
- Q. Where do you see application integration and open source heading over the next ten years, and what solutions will Talend have in place to accommodate that evolution?
A. Open source is the winner when it comes to innovation, and it will continue to lead the major advances of the next decade. But open source itself has changed in the past 5 years, being more open, more collaborative between vendors and user organizations than ever before (look at the Hadoop project, for example). As application and data integration evolve toward a unified approach and platform, like what Talend supports, open source will play an increasingly important role in that innovation thread as well.
Yves de Montcheuil is the Vice President of Marketing at Talend, the recognized leader in open source integration. Yves holds a master’s degree in electrical engineering and computer science and has 20 years of experience in software product management, product marketing and corporate marketing. He is also a presenter, author, blogger, social media enthusiast, and can be followed on Twitter: @ydemontcheuil.
APPLICATION INTEGRATION, DATA PRIVACY, Fresh Ink