Untangling SAP for Serverless Deployment

There are few technology trends that are a better fit for the current moment than serverless computing – particularly with its benefits of greater scalability, faster development, more efficient deployment, and lower cost. 

But it can be difficult to harness the power of serverless when critical data and workflows are housed in the powerful, widely used enterprise resource planning (ERP) system SAP. Organizations typically rely on SAP to centralize data management and optimize critical business functions like accounting, financials, human capital management, enterprise performance management, etc.

While SAP offers organizations a lot of flexibility, this means the SAP ERP is often complex. It typically consists of various disparate technologies, including a mix of legacy on-premise ERPs, S/4HANA, proprietary language, logic, business processes, and a myriad of integration tools that help connect and navigate this intricate landscape.

This complexity is especially challenging for organizations that are looking to leverage data and business logic in SAP to build digital services –  especially serverless functions.  

The Challenge

SAP provides a variety of integration methods and interfaces, which are supposed to help organizations connect and manipulate data and processes. The trouble is, rather than using industry standards, SAP has developed its own capabilities like:

• BAPIs (Business APIs)

• IDocs (Intermediate documents)

• ABAP (Advanced Business Application Programming)

• RFC (Remote Function Call)

• JCo (Java Connector)

SAP’s proprietary integration tools don’t extend well to third-party systems. Rather than facilitating integration, they complicate the process further. Let’s take IDocs as an example — there are over 600 IDoc types that you need to consider when developing a strategy for your organization’s integration project.

Add to that the fact that BAPIs don’t support all available SAP transactions, and that RFC might not be documented or supported by SAP, and the challenge becomes even more daunting.

How to Do It

Organizations that rely heavily on SAP can absolutely leverage the benefits of serverless. But it requires them to rethink conventional wisdom about integration with SAP in three ways:

Step #1: Bypass the middleware. Many organizations choose to overcome the hurdles of integration and SAP interoperability with middleware (e.g., ESBs) – including perhaps the best-known SAP middleware, SAP’s own Netweaver process integration suite. While Netweaver and other similar tools help to overcome common integration challenges, they tend to introduce their own complexities and pains – including an inflexible/monolithic architecture, complexity, and maintenance costs. Additionally, organizations have to install and support integration middleware that would allow the integration infrastructure to cope with upgrades and future changes to functionality — on both SAP systems and non-SAP applications. 

The best approach is to bypass the middleware altogether and leverage API connectors that can pull directly from the SAP system, expose BAPIs and RFCs, and parse metadata. The pros include faster implementation time (with less specialized expertise required), more flexibility, and lower maintenance cost. 

Step #2: Expose SAP data and business logic in a reusable way that digital developers can build on. Digital developers tasked with building, testing, and deploying serverless functions rarely have the expertise in SAP needed to quickly harness assets from the system. For that reason, any integration approach must include the ability to automatically generate a standard REST API and/or OData API and angular HTML5 from SAP. The automatic generation of reusable APIs – as opposed to point-to-point SAP system integrations – will lower maintenance and costs. And representing data and business logic from SAP in a standard digital format ensures ease of serverless development and testing. 

Step #3: Use a lightweight node.js implementation to bypass the cold start problem. Any team that has launched serverless functions (even if they don’t rely on SAP) is familiar with the cold start challenge. But the challenges of SAP integration can exacerbate the slow initialization times that often serve as a barrier to adoption of functions-as-a-service (FaaS). For example, the Java implementations and libraries that are common amongst middleware contribute to a large function size and slow start time. How do organizations jump this hurdle? In addition to bypassing the middleware (see Step #1), the single biggest breakthrough that SAP shops can use is to rely on a lightweight Javascript implementation for serverless functions. 

SAP is an immensely powerful ERP that is critical to the modern enterprise, but can present daunting challenges when trying to reap the benefits of serverless computing.

These are exactly the challenges we set out to solve with OpenLegacy Serverless, the first serverless-native hybrid integration platform built for legacy systems. OpenLegacy Serverless offers organizations running SAP a low-code environment to rapidly build, deploy, and run serverless APIs. 

Ultimately, modernizing their approach to digital integration with stored procedures can help these organizations rapidly accelerate function-as-a-service (FaaS) adoption within their organizations.