S/4HANA projects and their training data requirements

Testing testing

Thanks to our partnership with Worksoft, I’ve spent a lot of time over the last few months listening to testing professionals and trying to understand more about the challenges they’ve had around SAP for the last ten to twenty years. S/4HANA projects come in all shapes and sizes, but however organisations are planning to get there – RISE with SAP or ‘lie low’ and keep your perpetual licenses – there seems to be a huge drive from the top to step up the automated testing before and during the project, as well as making sure that when they go live on S/4HANA, the automated testing is already in place.

There are so many ways our Data Sync Manager™ (DSM) Suite can help with identifying and extracting information to build the test record sets (using the Object Extractor engines) and provisioning the latest versions of crucial master data that changes frequently (such as Condition Records and Material Masters) using Object Sync. Client Sync can also be used to refresh incredibly lean testing clients with just the latest master data and customising from Production, and even limited to specific company codes.

While most organisations are in the throes of planning or beginning their S/4HANA and/or RISE with SAP journey, there are some looking towards, or embarking upon, the end of the project – and they have significant training needs.

The training arriving on the platform…

This need to train at scale reminded me of a few projects I did with DSM over a decade ago when SAP was going through similar mass upgrades to ECC6. At that stage, many companies were looking to get accurate data but in more agile systems/clients, so they could provide training for far more end-users than they would normally do.

In this blog, I’m describing two particular examples that happened at two distinct UK customers. For both, there is a ‘don’t try this at home’ warning unless you are particularly well-versed in the capabilities of the particular DSM engines. Our Global Product Delivery team is on standby to assist when you mention my name and this blog.

Use case 1: It’s Client Sync, but not as we know it

The first story is from an organisation that had 24-hour shifts in place, and the end-users were to be trained during their normal shift patterns. It hit my desk (probably from Ikea but before the sitting/standing variations, because it was the late noughties) because the requirement didn’t seem something we could actually achieve. The training golden client was around 500GB (production being 2TB) and had a lot of master data and a year's worth of transactional history. The users were to take some existing transactions and master data and process them, as well as create some new transactions. We settled on the idea of multiple child clients (I think it was three but it might have been two – I’ve had two more kids and a dog since this project so my memory is not what it was) which were to be reset from the golden client, and a requirement was that the golden client could be updated but within very narrow parameters.

The parent and child clients for training were on the same system, and there was literally not enough time between training courses to be able to do complete resets with Client Sync. After some ‘interesting’ conversations with the SI, we agreed to a backup of the system before the training project started and leaving the system in ‘no logging’ mode. So if there was a fatal error, they would restore from the ‘back up’ before the project, and any delta changes in the golden client would be lost.

We then embarked on an activity of analysing the training scripts and linking the data that changed to specific tables, and also reviewing the biggest tables in the golden client by looking at Estimation runs within Client Sync.

The idea was that of the 60,000 plus tables in the training client, 59,950 or so were not actually changing as a result of the training. So – and this is the critical caveat the client had to sign off on – as long as no one went off script in the training, we could set a lot of the biggest tables to actually be excluded from Client Sync completely; i.e. wouldn’t be deleted, extracted, or inserted.

This is how we were able to set up an automated, scheduled set of Client Syncs to reset training clients from a golden client within a couple of hours. And this allowed the shift workers to be trained out of hours, without anyone needing to actually do anything to reset the data. Warning: If we got the scope wrong or someone went ‘off piste’, this would have gone horribly wrong, and probably while I and the IT team were asleep. It needs careful preparation and execution, and lots of sign-off from the business – but in this case was a game changer.

Use case 2: It’s not a PICNIC

A ‘PICNIC issue’ is a rather derogatory term (that I hasten to add would never have come from an EPI-USE Labs professional, but something I learned from a consultant at one of the big SIs back in the day – think Clinton in the White House and Blair in Downing Street).

PICNIC stood for ‘Problem In Chair Not In Computer’ – and it applied to me quite a lot of the time as I made my name as a particularly terrible/green Basis consultant. PICNIC issues though are the absolute enemy of IT trainers. One person in a class of 20+ people drains valuable time insisting ‘it doesn’t work on my computer’ and ‘it must be my system/data/luck’ (delete as appropriate or vary through the training).

With a little bit of discussion, and review of the training scope, we realised that by taking example Purchase Orders and using the cloning capability in Object Sync we would be able to make literally hundreds of completely identical data sets. We then went a step further and cloned the relevant vendor x times first and then tweaked the mapping table in DSM so that each copy of the Purchase Order made use of a different (but identical) vendor master. So everyone on the course could have their own vendor and purchase order to process, but since they were all identical, there was no doubt where the problem was if one of the trainees kept raising their hand.

To summarise: DSM provides incredible data agility for your S/4HANA projects or other testing/training needs. Reach out if you have a very specific need and see if our excellent consultants can provide a very bespoke solution.