Top Five Roadblocks to Cloud Integration by Sunil Mehra, Vice President, Middleware, Oracle India

Friday, 18 September 2015 08:18

In the digital economy, integration is no longer a technology term—it’s a business imperative, particularly in ASEAN, which is experiencing a huge acceleration in the corporate cloud adoption..

Now, information must not only move across myriad on-premises systems, but also across mobile and cloud-based platforms and apps in order to support these digital business models.

Unfortunately, integration challenges consistently hinder digital transformation projects, particularly when it comes to the cloud.


For example, quick access to cloud capabilities has given rise to so-called ‘shadow IT’ projects, in which line-of-business (LOB) leaders bypass IT to sign directly with cloud vendors for function- or department-specific applications.


Such projects can reduce time to market and increase agility in the short term.


But when LOB workers want to augment the application with information from existing systems maintained by IT, integration becomes a potentially huge problem, particularly if the cloud-based application does not adhere to IT’s standards.

Moreover, many companies today do not have an enterprise-wide cloud investment policy.


Instead, they choose multiple cloud vendors to get the advantage of best-of-breed technologies.


However, history has shown that sooner or later, companies will want these systems to share valuable data, and once again integration becomes a roadblock.

As companies add a new level of cloud-based software and systems to on-premises infrastructures, many have grown frustrated with increasingly complex integration requirements.


In fact, a recent study found that one out of every two companies has abandoned its cloud software as a service (SaaS) applications in the last three years due to integration problems.

To overcome that challenge and to build a harmonious hybrid environment, IT and business leaders must collaborate on an integration roadmap that covers five critical aspects:

1.) Data Mapping. In order to share and integrate information, companies need to discover and map data locations throughout the organization, from isolated departmental SaaS apps to enterprise ERP systems running financials, supply chain, and other back-end functions. Known as data mapping, the process is often viewed as one of the most important and complex tasks in many integration projects. Cloud-based integration platforms like Oracle Integration Cloud Service offer pre-built tools and connectors designed to help companies simplify the process of integrating cloud services as well as on-premises applications.


2.) Data Governance. Taking a unified approach to data governance is vital to maintaining data integrity, particularly as many companies want to maintain hybrid environment for their data and applications. As a result, lines of business and IT need to agree on and implement a set of policies around data governance that clearly delineate when and how companies will integrate cloud resources, both to each other and to on-premises applications.


3.) API limitations. Application programming interfaces (API) are sets of routines, protocols, and tools for building software. However, each SaaS vendor and cloud provider enforces unique API frameworks, security standards, session management practices, and maintenance procedures, and what they provide varies widely. For example, API tools in many niche SaaS apps aren’t built for full integration—they may not be able to expose the full data model, or have limited operations. Because of this, organizations need to analyze the interfaces offered by their cloud vendors and plan according. For many, this means adding a centralized layer that sits on top of the APIs to support full data integration.


4.) Security. The increased access points and data integration requirements of a hybrid environment can represent security vulnerabilities. To start with, the underlying platform must have strong security layers embedded throughout the stack. On top of that, any integration services and tools must support security policies used by many applications, such as HTTP Basic- or WSSecurity-based authentication.

 

5.) Talent Shortage. Finding the people with the skills to implement integration tools is also a problem, and many companies don’t have the right talent in house. One step to solving the skills shortage is to use cloud integration services that cut down on the workload via ready-built connectors for many applications, and a development environment requiring minimal configuration and zero coding—cutting down the need for developer expertise.


Planning for Cloud


While the integration discussion sounds technical, the ramifications are not.


Indeed, many companies have found that a poorly implemented integration strategy can trigger significant business challenges, such as the inability to react to unified data in real time.


Companies need to think about long term issues and adopt the appropriate cloud integration technologies because in today’s always-on business climate, information access can make the difference between success and failure.

Last modified on Friday, 18 September 2015 08:30