Typical reasons why customers are not doing housekeeping
Everyone knows housekeeping should be done, but according to our research in over 150 SAP BW systems, only a low percentage of customers are housekeeping their SAP systems to its fullest extent. These are the typical reasons for that:
1. UNCLEAR RESPONSIBILITIES "Know one knows who is responsible"
Take statistics tables as an example, in which SAP BW creates vast amounts of data, the responsibility for many of these tables is spread between the Basis and Applications and Operations team. Ultimately whose responsibility for these tables can be unclear as there is potential for all these teams to be assigned for the cleanup of the tables. The simple approach would be to create housekeeping rules jointly; therefore input is needed from all involved parties. Assign the responsibility for the running of the tasks to one team as part of their regular activities. This will ensure clarity as to which team is accountable and should make time, regardless of other projects / tasks that are deemed a higher priority. Which brings us to the next point.
2. LOW VISIBILITY IN THE ORGANISATION "Customers don't know how much potential they have"
Over other projects housekeeping will always have a “low priority” as the end user of the system who are mostly running queries, often the view is taken that housekeeping does not directly impact or generate value for the end-user. However, the best way to gain visibility and lead to a housekeeping project is to gather the data, and provide projections as to consequences, both performance wise but mostly cost wise for not having a housekept system. In our analysis we often see housekeeping potential of 30 % of the system size! That is quite some costs.
3. NO RESOURCES "I have to time for this"
Cleaning up is usually very resource intensive, so even with a cost saving the business case for cleaning up the system can be not good enough, especially when good housekeeping practices should be done on a regular basis. This can often be a catch 22 situation because at the same time if no clean up is performed, this will lead to an unnecessary increase in system size and can impact the performance of the system.
Now we’re going explore some of the more technical reasons that stand in the way of housekeeping an SAP (BW) system. Earlier in the post we mention that the housekeeping can be resource intensive. One of the main reasons for this is the housekeeping tools provided as standard are spread out within the system. To show an example of what I am referring to, click on the this link. This is a list provided by SAP for a before HANA migration system cleanup for a SAP BW system. As you can see, all the tasks use their own program that needs to be started individually and would need to run in every system. In addition, if you are not familiar with the program, and unsure as to what data will be affected or what the expected runtime, in this situation I am sure you will be very hesitant to run it. For such cases, a central repository of documentation would help give good a indication, or expectation as to what the tasks do, and recommendations for running the task.
5. NOT ENOUGH FUNCTIONALITY "I would delete it but I cannot"
Our analysis often shows large PSA and ChangeLog tables. Our general recommendation is of course: DELETE IT! (at least all data older than 14 days) But, deleting all data in PSA and ChangeLogs is not always possible because the source system may be too unreliable, or the data model requires you to keep the data. For most systems there is no near-line storage for archiving in place, which allows for moving the data to a write-optimized DSO and “nearlinestoraging” old data, so limiting the available options for the data to (1) keep it or (2) delete it. This dilemma applies for many areas, where the only options available are keep it and delete it. Not all data can be deleted, but you cannot just keep all data. If the data is kept, a solution to accommodate this would be to compress the data and minimize the storage space required, but is it feasible to just keep all data, especially the temporary type, that offers little or no value.
6. TOO HIGH RISK "Business does not want to delete any data"
A combination of factors, whether it is budgetary or functionality may lead to lack storage space, a point where the only option is to delete. If forced down this path with limited understanding will become a high-risk scenario because how can you ever be sure, you are not deleting too much? The best way to mitigate this is to become proactive to build up your understanding of how to housekeep your systems rather than take the reactive approach to housekeeping. A line needs to be drawn and strategy made as to what to keep and what/ when to delete.
7. NO TRANSPARENCY IN THE SYSTEM "I have no clue if housekeeping is done and to which extend - there is no documentation available"
The final point, as you’ve seen, the standard tools for housekeeping, factor in that housekeeping should be performed system by system and on a regular basis. Unfortunately there is no central space where housekeeping can be trackedand progress monitored provides no transparency. Central management for housekeeping is the key for an increase transparency, reduce effort drivers, for running all tasks and reduce the risk of deleting data that should not have been.
Many SAP tables contain temporary data, data that lose value quickly after its creation.
Often these tables growth rapidly because data after the end of purpose is not handled properly and regularly.
Sometimes because people are reluctant to delete any data.