I just took my trusty wheeled steed in to have a 100,000 mile service and cam/timing belt replacement and while checking it in of course I was told that the belt should have been changed at 90,000 miles. So much for internet reviews of timing belt appropriate interval advice. I am pretty sure that 100k was ok, though it is possibly on the high end of the risk graph for failure. Nonetheless, the visit was a bit like answering the question at the dentist as to how frequently and how one flosses ones' teeth. Invariably my answer still results in a chastisement and perhaps a quick lesson on how to do it, why to do it and how frequently. Dentists would have us flossing our teeth several times a day and auto mechanics would have us getting oil changes and services every 2500 miles if they could convince us that the benefits outweigh the costs. The walk home through the budding leafy green streets of my neighborhood in the seasonal drizzle got me to thinking how easy it is to translate this experience to running SAP systems. Early on in my SAP career I had the distinct pleasure of attending a working session on archiving on SAP systems and there was a loty of emphasis on getting an archiving strategy in place as soon as possible to avoid a legacy of pain in the future. As many of you will know, archiving is more or less akin to flossing. Uncomfortable, difficult but ultimately necessary.
No-one in their right minds would openly say, I don't floss, and my teeth are full of scale and plaque. SAP system owners also never say, we don't archive, our database is growing at a rate of xMb per day and the SAP disc footprint increases by x% per annum, or at least if they do, they shouldn't be proud of it. Yes, hard disc technology is very cheap these days and one can fill disc with as much stuff as one can, but is this ultimately helping the cause? The answer has to be an emphatic no. If you run a patient administration system then fair enough, having a lifelong history for a patient is indeed useful information. You wouldn't be happy if every time you went to visit the doctor or dentist, it was like dealing with a Dory, as in the Disney-Pixar animated film Finding Nemo - every visit would be excruciating as you would have to spend loads of time explaining your history or having your teeth counted.
Hardly anyone can argue though, that there is a strong business case for archiving, not only to save ultimately on operational costs, backup time etc but also to save the business from itself. Having thousands of pricing condition records for example, whilst impressive, is less impressive when you consider the performance hit that this brings to bear on your system. The legacy of more than 7 years of active financial data is unnecessary from an audit and compliance perspective and besides, archiving means just that, and is not the same as deleting. In my current and past roles the idea of archiving comes up very frequently. Those of us with type A personalities can relate to the ‘control' aspect of having all our ‘stuff' in one place and not having to compromise on having it sprinkled all over the place, which is something that the archiving notion suggests will happen. In the days of yore, archiving invariably meant dumping data to magtape and then having to beg with an IT person to restore it to some locale if we needed to look something up. Modern day systems are a little more elegant, you may lose some of the granularity but in the grand scheme of things, these are relatively easy pills to swallow even for the business.
As an individual in the workplace we of course have a personal responsibility to manage and archive the data that we are personally responsible for. We may have no control over the over arching strategy that the business has with respect to SAP system archiving but we can be advocates for and should be advocates for archiving in its most fundamental form. A good starting point is asking our IT people hard questions, like - "do we have an archiving strategy?" If the answer is yes, then the next question would be, how frequently we exercise the practice of the strategy and whether additional archiving objects are evaluated periodically. Every new initiative should have archiving related questions incorporated into the feasibility analysis, the design and the supportability design. A simple example would be something like internet sales. Every shopping cart creates a record somewhere, even if that shopping cart is abandoned and the person walks away. What happens to that abandoned cart, is it still resident in the system somewhere? Every cart that runs a full lifecycle, becomes a sales order and that order is settled and delivered against, should ultimately be rolled up and archived. You should check with your auditors as to what the recommended policy is and should be for sales order record retention, and if your organization doesn't have one, push for one to be established, remember too that the sales record may be resident in more than one system and your archiving efforts will probably need to span multiple systems.
One of the easiest ways to determine just how old some of your data is, of course is the running of reports out of your SAP system. Those can be time consuming and a drag on system resources. Another option worth considering is single table inspection at the transaction level. While structures are difficult to read with query tools like SE16, SQ00 and SQVI, most sales order information at least can be inspected at some level in ECC, with these and 3rd party query tools simply by looking at VBAK.ERDAT as illustrated here with some basic data criteria.
Other tools that you might want to consider, are simple things like using transaction VA03 to look at your earliest of online sales orders, while not very scientific, taking an early order number can help you understand just how much irrelevant junk you might have still lurking in your system and come to some conclusions as whether it is time to clean house.
