Clean Core Measurement - Overview
A key aspect for transparency and motivation is the measurability of Clean Core and the progress within the S/4HANA project. This ensures that the project's status is visible and traceable at all times.
Table of contents
In this article, we'll introduce the topic of measurability, examine its importance, and then discuss current possibilities.
Introduction
If you're currently working on a clean core project, there are actually three variations, to be precise: a greenfield project, a bluefield project, and a brownfield project. With a greenfield project, you build everything from scratch, giving you the opportunity to implement clean core standards. Functions are checked beforehand to see if they are included in the standard, and extensions should only add value. With a bluefield project, you adopt components from a legacy system and modernize them piece by piece at a later date. With brownfield transformations, you essentially take an existing system, upgrade it, and then examine what actually needs to be corrected to get the system clean. Further information on the different types can be found in the ABAP Guide from DSAG (soon also available in English).
It is therefore crucial to have transparency regarding error messages and critical objects in the system in order to correct them. Measurability plays an important role in all three types, because we want to keep our system consistently clean. This means we have a governance process for the system, or we want to get our system clean. This means we look at which findings we should address first and what is currently most critical in the system. Measurability helps in this regard, as it transparently presents Level D objects and defines a Technical Debt Score for in-house developments. This gives us various options for addressing the findings. Whether it's first addressing all Level D objects or modernizing applications with a high score, we have the opportunity to tackle the challenge transparently.
Goal
There are different types of goals for companies. Clean Core is one of them. Achieving this is now relatively easy thanks to the Clean Core Level Concept. Basically, all we need to do is remove all Level D findings from the system and continuously work on Level C.
With that, we'd essentially have a clean core and be upgrade-stable. However, becoming cloud-ready requires a much greater effort, as we need to achieve 100% Level A. We can then calculate corresponding values for these two metrics by, for example, taking the objects from the different levels and calculating a percentage. Theoretically, this would allow us to always provide our management with a KPI that reflects the current state of the system.
Challenges
However, there are also some challenges we should consider. For example, the system wasn't modified just for fun, but because we want to use it to map specific business processes. This means that modifications or enhancement technologies, which may be outdated, are nevertheless necessary for us to keep our processes running smoothly within the company and to generate added value. The second issue is the lack of APIs. Sometimes we use Level C APIs because there are no alternatives or approved APIs in the system, such as ABAP Cloud or Classic APIs. This means our debt in the Level C area can increase relatively quickly, especially if we are working in modules that have not yet been enabled for ABAP Cloud.
Measurability
Therefore, let's look at two current projects regarding measurability and what we can already gain from them today.
RISE Dashboard
The colloquially known RISE Methodology Dashboard is a standard SAP product which, as the name suggests, is primarily intended for RISE customers. It provides us with a standard dashboard where most of the data is already pre-configured and which is set up in a standardized way to measure our systems and obtain key performance indicators. However, and this is the biggest drawback, we need a RISE contract with SAP for this, meaning we have to be on the private cloud. Since many customers are not yet on the private cloud or don't want to be, the dashboard is only available to a limited number of companies, and a different solution is needed to make the measurability of Clean Core visible. The dashboard is currently divided into the dimensions of Extensibility, Data, and Integration and is intended to provide figures and overviews for these dimensions. Looking at the Extensibility, we obtain the Technical Debt Score of a system, as well as key figures for the modifications.
In the lower section, we receive key figures for the different levels of objects. Objects for Level A are automatically determined from the system and stored. All objects from standard ABAP for Levels B to D must be imported via an ATC check. We then run a check using the Custom Code Migration App or a Central ATC and upload the result to the dashboard.
From here, navigation to the various findings and objects (Levels B to D) is also possible to view findings and messages for individual objects. This allows Level D objects or objects with a high score to be addressed first.
Project Kernseife
Project Kernseife was also launched here, which further refines the measurability of Clean Core. This is especially important when it comes to documenting unreleased APIs that are nevertheless useful, such as Core Data Services that are already in the system but do not have a release contract. Furthermore, the number of ATC findings output at the end of the ABAP Test Cockpit check should be reduced to focus on the essentials. To this end, various CAP applications were built within the project, which the customer can install on Cloud Foundry in the BTP to maintain the exception list for ATC and define customer-specific successors. The project also provides a customized ATC check to allow the defined exceptions to be used within the customer's own system.
However, this solution is not yet 100% cloud-compatible. This means it does not work in ABAP Cloud and therefore cannot be run on any central ATC within the ABAP environment. Furthermore, CAP applications need to be deployed and operated, which may not fit the current architecture of the company. More information on measurability with "Kernseife" you can find it in the link collection below in the SAP Community and in the GitHub project.
CCM
For this reason, the CCM project was created to address various points in the chain and also to give non-RISE customers a way to easily implement measurability.
Reason
Basically, the two solutions above, namely the RISE dashboard and the "Kernseife" project, already work very well if you want to achieve measurability within a Clean Core and customize it to your own company. However, the first solution lacks components such as documentation and assignment to products within an organization or the formation of clusters to gain an overview. Furthermore, customers also want to document when they use specific APIs in the system, but these are actually needed to advance the process. We don't want to see such findings in the Clean Core dashboard every time, but rather document them once and then be able to extract them from the findings accordingly. Besides this hypothetical Clean Core state, we naturally want to regularly check how things compare to the actual standard. That means, if the goal is to move to the public cloud, we naturally want to see which processes we haven't implemented yet and which are still open. We always have a working version and an actual state in the system, which we can then report on accordingly. Finally, the process should also be automated so that figures are automatically made available based on weeks or months.
Requirements
We have currently set the following points as requirements to develop a Clean Core Measurement that can offer us added value:
- Open Source - The whole thing should later be freely available as an open-source project so that anyone can install it and be supported on their journey towards a Clean Core.
- Central ATC - The solution should run directly on the central ATC system. This gives us the opportunity to automate the runs and imports for the objects, so we no longer have to think about implementation and always receive up-to-date key figures from the system. Furthermore, possible custom code checks should later be able to be implemented directly in the current system with the data.
- Documentation - Documenting findings, packages, and tasks should be easy and integrated with the dashboard. This should allow for cluster assignment and documentation, as well as adjustment of the custom score to keep the to-dos visible in the system.
- KPIs - We want to see all systems at a glance in the dashboard and be able to view the two KPIs for Clean Core, Upgrade-stable, and Cloud-ready at any time.
- Reporting - Through reporting, we want to gain further insights into various levers we have in the system and how we can implement them. We want to evaluate findings by group and message, but also by the different clusters found in the system.
Outlook
Work on the CCM has already begun, and in the next articles we will give you insights into the architecture and the decisions made, so you can get an overview of the project. There is already a GitHub repository for the project, where you can always find the latest version of the solution. The repository will gradually fill with applications and tutorials.
Conclusion
Measurability of clean code remains an important component for achieving transparency in the Clean Core process and development. We want to improve the state of the system and be able to provide information at any time. It definitely makes sense to implement measurability here, regardless of the area or project in which it is ensured. A fundamentally negative approach is not to perform any measurement and to rely solely on intuition. Especially if you have a lot of developers working in the system.
More information:
SAP Help - SAP Cloud ALM and RISE with SAP Methodology
SAP Community - Project Kernseife - improving the measurement of Clean Core in ABAP
SAP Community - Taking Control of Clean Core: When to Use Project Kernseife in addition to Standard SAP Tools
GitHub - Project Kernseife
