OpenLegacy Blog

ETL vs. API: Understanding the Implications for Legacy Modernization

Share this:

diag-dataimportBusiness is changing, but there are two things you know for sure. You love your legacy mainframe systems, and you need to modernize to help keep your company competitive in today’s mobile world. Let’s face it, there are more important things to spend your money on than rewriting all of your legacy code so it is compatible with next-gen environments.

Data integration is one way to go if you want to keep your legacy systems in place. In this post, we’ll present a comparison of two potential data integration approaches: ETLs vs. APIs.

Extract, Transform, and Load:
Old school isn’t best in this case

Many want you to think that it’s time to upgrade your legacy systems, but there’s still so much they can accomplish in app-centric business. Just because your legacy systems still have life, doesn’t mean traditional data integration approaches are appropriate.

ETLs gained traction in the early days of data warehousing, but more for their ability to move data rather than integrate it in the sense that you need today. According to Anant Jhingran of Apigee:

Expensive ETL systems were necessary to de-dupe, cleanse and normalize [information]. However, because the nature of the data changed rarely (business systems were well understood, well modeled, and stood the test of time for decades), such investment was typically justified.

Essentially the goal of implementing ETLs was to consolidate data and move it between systems without investing in expensive storage systems. If this sounds different than the problems you’re experiencing with modernization, that’s because it is.

Although InetSoft says that ETLs are great for bulk data movement, they say that they “aren’t that great at near-real time or on-demand data access, where you have a need for quick response.” On-demand data access is precisely what you need for mobile business.

Your customers need to interact with you through many different apps and services, and Jhingran says, “the data that they generate, changes very rapidly and in many cases, is not even in the control of the enterprise that needs that data.” Even though you could add functionality to ETLs, the process would be expensive and inefficient in comparison to an API approach.

Application Program Interfaces:
New school is your key to legacy modernization

APIs can help you modernize the legacy systems you’ve come to love by correcting on all of the weaknesses of traditional ETLs. Instead of moving data and consolidating it to avoid costly storage upgrades, APIs are able to extend your previously coded data into modern business processes.

APIs enabled a fundamental shift in how data is handled. Where developers used to focus on storage optimization, they are now working to ensure data, applications and other IT assets can be consumed without interruption.

By using APIs, developers allow mobile, cloud and web applications to consume legacy assets without require any backend modifications.

Modernization made easy

In the end, the ETL vs. API showdown isn’t much of a contest in terms of modernization. ETLs are expensive, rigid, and ill designed for current consumption trends. APIs, on the other hand, allow you to “connect to a remote service to push files, retrieve data or manage computations. By harnessing an API you can deliver lightweight apps without losing the key functionality or service of your business,” according to V4 Technical.

With OpenLegacy’s API-centric approach to data integration, your legacy systems can be extended into the modern business world without shattering your budgets or forcing you to retrain entire departments.

How loyal are you to integration methodologies? Would you rather switch integration techniques than replatform? Let us know in the space below.


Tags: APIs, Legacy Modernization

Share this: