Loading...
You are here:  Home  >  Data Education  >  Data Architecture News, Articles, & Education  >  Data Architecture Blogs  >  Current Article

When to Use iPaaS vs. ETL

By   /  June 14, 2017  /  No Comments

Click to learn more about author Paul Varley.

With more applications around, usually from different vendors, the need for data integration has never been greater. New integration approaches have emerged, and today they coexist with other integration forms that have been around for years. Today we will compare an on-premise ETL tool with an Integration Platform as a Service (iPaaS) and answer the question, “When would you use one or the other to address your integration needs?”

When to Use ETL

As the name implies, Extract Transform and Load (ETL) software manages three aspects of the integration process. Extraction involves getting a copy of data from a source, which could be an application, database or text file. Transformation translates the source data to match the format of the target system. This includes changing data types, combining or splitting fields, and applying more complex formulas. Loading completes the process by putting the transformed data into a target system.

Many companies use ETL software to load data from different systems into a Data Warehouse for reporting and data analytics. It moves data in batches, often on an hourly or daily basis, so it works well for moving data behind a firewall inside the organization when the data is not time-sensitive. The drawbacks of ETL are that it is not well suited to real-time processes or more than two connections at a time. You provide the server, which means your IT department gets involved, and most ETL tools are sold as a software license plus annual maintenance, so the up-front cost is usually higher than a Cloud service.

When to use iPaaS

An Integration Platform as a Service (iPaaS) provides integration capabilities as a multi-tenant Cloud service. Often an iPaaS is used to replace an ETL tool because an iPaaS also allows you to extract, transform, and load data. An iPaaS, however, also allows you to build and deploy even complex integration projects with more than two connections and even event-driven workflows. The user interface of an iPaaS often has a shorter learning curve. Unlike more traditional data integration platforms that are associated with slow processes, an iPaaS allows you to create integrations in a short time with a web-based interface that you can access wherever you are. An iPaaS will include connectors to the main SaaS applications as well as technology connectors for common databases, text files and protocols such as OData.

If you’re integrating Cloud applications, you will not need any server resources with an iPaaS, but an iPaaS is not just for Cloud-to-Cloud integration. It also handles integration with on-premises systems, usually by installing agent software behind the corporate firewall. Like most SaaS applications, an iPaaS is a subscription service with pricing based on the capabilities the customer needs, which allows you to pay as you grow into the service instead of incurring a big up-front expense for a software license.

Which approach should you take? If your company already has an ETL software license and IT staff with the skills and bandwidth go get a new integration up and running in your timeframe, ETL may be the best fit. If any of the following describe your situation, choose an iPaaS:

  • You do not already have ETL software
  • Your users are business analysts
  • You require real-time integration
  • You need integration between more than 2 endpoints in a single map
  • Limited server resources available

About the author

Paul Varley, Product Marketing Manager, Scribe Software. Paul is a Product Marketing Manager at Scribe Software. Paul helps enterprises and systems integrators use the Scribe integration platform as a service (iPaaS) to accelerate their application and data integration projects. Paul has more than 20 years of product management and product marketing experience and has managed numerous enterprise software and networking products, most recently with Arbor Networks. Paul earned a Bachelor of Arts in Computer Science from Dartmouth College and a Master of Science in Information Networking from Carnegie Mellon University. Follow Paul and Scribe Software at: Facebook, LinkedIn, Twitter, YouTube

You might also like...

Is Data Governance Solely About Controls on Data?

Read More →