Moving data between systems? Think DSM
If you have never worked with DSM before, read on – it could just save your project! If you’re considering a system split, or carve-out, or there’s a merger or acquisition looming, see what DSM offers before settling on any alternative approach.
If it were a simple matter of uploading data to Production, many other tools could do the job. But the projects that we’re talking about here are complex: the original data needs to be converted, or carved out by careful selection, or manipulated according to business rules, or any combination of such requirements. Normally, this translates into extensive custom development or lots of manual activity – not to mention late nights and working weekends – but DSM can automate such requirements and produce repeatable, predictable results. High performance means that you will have enough time to test and re-test various scenarios thoroughly before the big moment arrives.
EPI-USE Labs has successfully delivered many such projects globally over the years, and has gained valuable experience and feedback on each. The following paragraphs will share some of these experiences with you, and you might just find that your requirements are not in uncharted territory after all.
DSM comprises four independent yet complementary components:
System Builder: quickly create a new system shell, minus any application data. This is done by making a complete copy of all repository objects and client-independent data in Production, then using that to build the new system.
Client Sync: copy a subset of Production data to create a smaller target client. Specify your subset by leaving out older data, or by excluding unnecessary large tables (e.g. logs), or by selecting only parts of the enterprise based on company code.
Object Sync: create reliable test data yourself as and when needed – from a single object up to a whole range depending on what fits the scenario best. Vendors, employees, materials, business partners – they’re all there: supported, automated, secure. Data can be scrambled, converted and/or cloned during the copy.
Data Secure: mask sensitive data in-place on a test system, regardless of how the data was created. The most common scrambling needs are delivered as standard options, but you can add your own rules too. Data Secure has been optimised for mass operations on large data volumes.
Let’s look at three successful projects, all transformed by DSM.
DSM in a split
A system-split or ‘carve out’ as it is often referred to, involves extracting only selected parts of a system to create a new one. This typically occurs when part of a company is sold, or if a company wants to decouple HCM from the rest of ERP.
In DSM terms, System Builder extracts the repository, cross-client customising and default client data (client 000, 001 and 066) from the source system. Then, once the new system is installed, Client Sync’s enterprise slicing capability selects the master and transactional data, plus their dependent enterprise keys, by company code and copies this data to the new system.
Example #1: European construction-services business
A well-established global construction company sold off a portion of its business earlier this year. In functional terms, this meant doing a complete company carve-out (in this case for 5 company codes) of the data across multiple SAP modules: FI, CO, MM, SD and HR. DSM System Builder was used to create the new system; then Client Sync was used to do the enterprise slice carve-out for all modules except HR; and finally Object Sync was used to copy over the HR data.
- 10 years’ worth of data on source system, about 4.5TB in size
- Live cut-over was done in one day
DSM in an HCM merger
SAP system mergers occur when two companies join, or an internal restructuring is carried out, e.g. to consolidate disparate systems from multiple countries into one global system. The main challenge in these projects is that the enterprise structures will probably include clashes, and much of the basic customising will have to be redone. Identifying and resolving these clashes requires an enormous amount of work and the new configuration must be completed before the data migration can occur. DSM fulfills the dual role of converting and migrating the data.
All DSM engines can transfer data between unicode and non-unicode systems, as well as across differing operating systems (endianness). Object Sync includes upgrade logic, which means that it can copy data from a lower to a higher SAP release, or from a lower enhancement/support pack level to a higher one.
Example #2: Barclays (South Africa & United Kingdom)
This massive data migration project was heading for trouble until DSM came to the rescue. Project milestones were slipping dangerously, one after another. The chosen technology, LSMWs, simply could not handle the complexities. With little over a month to go, EPI-USE Labs consultants came in, built and tested all the required conversions, and migrated the data. The live cut-over was accomplished in a weekend and the go-live date was met. Some project statistics:
- Total number of employees: 41 000
- Total export time: 1 hour, 6 minutes
- Total import time: 1 hour, 57 minutes
- Number of functional conversions: 120+
If it wasn’t for DSM we wouldn’t have been live on the 1st April. What we achieved in that short amount of time is amazing!
– Francois van Wyk, Program Manager, Project Nexus
Example #3: British Petroleum (United Kingdom)
Over the course of two release cycles the following data was migrated from one client to another in the same productive system:
- 3778 employees migrated between production clients
- 22 945 changes to records using 14 custom conversions
- Total export time: 12 minutes, 45 seconds
- Total import time: 4 minutes, 8 seconds
DSM, and the highly competent EPI-USE DSM consultants, facilitated the successful migration of the SAP data of multiple countries from legacy systems onto one central system. We would gladly recommend EPI-USE DSM and EPI-USE consultants to anyone.
– Silvio Lanaro, BP Vice-President of HR IT&S
For more information please contact firstname.lastname@example.org