For better or worse, cloud services such as Dropbox and Evernote are here to stay, and admins better have solid cloud application integration systems if they don't want to lose
Cloud application integration systems are the approaches your organization will take to put existing applications and IT infrastructure together with the personal cloud services your employees use to conduct business. If your organization embraces these services, be proactive in integrating them before apps and data get compromised.
Here's one example of why cloud data integration is important: Employees might store Office documents in personal Dropbox accounts and collaborate with others on those documents via the corporate content management system. In such circumstances, it's just a matter of time before files become out of sync, changes get overwritten and efforts are duplicated.
Comprehensive cloud application integration systems helps prevent these types of scenarios, but you'll have to take a number of considerations into account, including which services to integrate and how to go about it.
Types of cloud data integration
The most common approaches to cloud data integration fall into four categories:
- In-house development: The organization develops, implements and supports its own cloud data integration system using specific processes, proprietary application program interfaces (APIs) and monitoring tools, building in redundancy and supporting consumer systems that are constantly changing.
- Vendor outsourcing: The organization hires a consultant to do the same work IT would under the in-house development option. Vendors may also support their cloud application integration systems once they're implemented.
- Out-of-the-box: The organization implements prepackaged cloud data integration software, either in-house or through a hosting service. Development usually isn't involved, but the company has to purchase and license the software, which can be costly. IT also has to install, configure and maintain the software, which can drain its resources.
- Software as a Service: The organization uses a cloud-based service for its integration needs. As a result, IT doesn't have to be concerned with many of the implementation and maintenance issues of other options, but it does mean giving up more control.
You might also find that a hybrid approach works. For example, you can integrate data in-house and use a cloud service to integrate applications.
Planning cloud application integration systems
More on cloud application integration systems
Keys to integrated cloud application management
Profiles in application integration
Cloud connectors gain traction in enterprise application integration
With these approaches in mind, start looking at your business requirements. Before you can choose and implement an integration system, you should have full knowledge of the business processes and workflows that will tie into or be affected. Take into account issues such as the direction corporate data flows, the number of apps that connect to the data and how often you need to sync that data.
Carefully distinguish between cloud data integration and application integration. The files workers move between Dropbox and the in-house wiki are one issue, and the applications employees use to access and modify those files is another. You need to understand both components to effectively integrate all the moving pieces.
The next issue: How will your in-house system communicate with consumer services? For example, if employees use Atlassian's Confluence to collaborate on documents and Evernote to capture information, you might be able to take advantage of the StiltSoft plug-in that integrates the two. But if workers use a proprietary system for collaboration, you'll have to determine what capabilities exist within your system and in Evernote to make communication possible.
Another consideration is data transformation (or data translation). Moving data between systems is often the trickiest part of an integration strategy. Simple data distribution, such as copying Word files between two systems, might not be a big issue, but data stored in different file types with different metadata structures is a more serious concern. Synchronizing data between systems often results in complex, resource-intensive solutions.
Resource considerations are not limited to data transformation. Factor in the resources that are currently available to you, plus the resources you'll need to implement and support all your cloud application integration system's components. This can include everything from the hardware you need to host data and applications to the personnel that keep everything running. If IT resources are already limited, an in-house integration system might not be possible without significant investment.
Other considerations you should take into account are related to the actual implementation. Issues such as performance, latency, scalability and availability are all tied to the type of integration you implement. For instance, cloud-based integration might have more latency than the in-house option, but the cloud might be easier to scale up and down without extensive hardware investments. The better you understand the performance requirements -- for now and for the future -- the more effectively you can move forward with cloud application integration systems.
Finally, consider application and data security. In most cases, you have little control over the data stored in a service such as Dropbox. But you can control the integration processes themselves. You might decide that your integration system will live behind the firewall to maximize control over that data. Before making any decisions, ask yourself how much control you're willing to give up.
This was first published in August 2012