Anyone involved with document management in an SAP context has always had two basic options: use the SAP Content Server or connect ECM/archive systems via ArchiveLink. Just as storage technologies have developed further over the past few years, these kinds of third-party systems with their myriad functions are increasingly becoming unnecessary. Document display and the administration of functional metadata can now take place entirely via SAP, while storage and media administration are transferred to a hierarchical storage management system optimized specifically for the purpose.
As a free in-house SAP product, the SAP Content Server stores documents on a separate instance alongside SAP, comparable with a content-addressed storage (CAS) system. It has, however, not been designed for the legally compliant administration of large quantities of documents during live operation. For this reason, the bulk of SAP users – according to KGS observations between 70 and 80 percent – are now looking at connecting ECM systems.
Classic archive systems constitute a “black box”
These kinds of classic archive systems (or DMS/ECM) were originally designed outside the SAP context for archiving any process-related documents. They therefore combine extensive components for archive management – far broader than SAP requires – and constitute a black box vis-à-vis SAP.
Who is responsible for access, metadata, and authorization?
The most taxing question is the issue of access and metadata administration – both are functions that the archive system claims for itself. Access to SAP-related documents via the archive is, however, unnecessary since the SAP-GUI already provides a fully fledged document viewer. In order to utilize the full archive scope, it is also necessary for the archive to know the metadata required for a search and for the authorization system to also be used. Here there are inevitable conflicts with SAP, which claims for itself metadata administration and the issue of authorization. Conclusion: classic archive, DMS, and ECM systems can only be used optimally if the built-in metadata administration, authorization logic, and the actual client are used.
A new archive approach tailored to the SAP archiving requirements is therefore necessary. This leads away from traditional SAP document archiving involving third-party systems and simply requires in its place middleware between SAP and hierarchical storage management (HSM). All of which necessitates changes to the storage systems which have evolved over the past few years.
Archive systems used jukeboxes for storage purposes
Originally archive systems would use optical and magneto-optical storage media for storage purposes, such as jukeboxes (systems that hold multiple optical data media). There were no generally valid drivers for these devices at the time. The upshot was that every archive system implemented its own driver and could not rely on standards such as file systems even for storage, but described the media in its own formats and storage logic.
How hierarchical storage management (HSM) systems pave the way for new archive concepts
State-of-the-art HSM systems by contrast act as a separate infrastructure component, separated from the central unit of server or archive system. They are able to administer disparate media (tape libraries or other storage classes such as SAN/NAS), provide standardized file system interfaces, and support replication scenarios as well as data migration. Virtually all SAP users are currently involved with the procurement of such HSM systems or have already introduced them. Since only leading applications such as SAP access HSM systems as the infrastructure component, these applications are also responsible for the entire authorization management. The same applies to the administration of the metadata required for the search; this data is administered solely via SAP as the leading application.
What the new archiving philosophy entails
This approach paves the way for a new archiving philosophy, based on the fact that every system only performs the task it was intended for and uses the systems optimized for the purpose for all other tasks. In the case of SAP archiving this entails support of the standardized interface ArchiveLink, the high-performance administration of the metadata required for storage and access, and the optimized addressing of the storage systems used for storage. In short, archive systems for SAP thus only need a simplified architecture and can dispense with a large part of the usual functions.
SAP is basically also designed exactly according to this model: for the ERP system an archive solely constitutes a data sink along the lines of an external data store. Access to “outsourced” data and documents takes place via a primary key, with all the metadata required for a search being administered within SAP as the leading system. Thus there is no need for SAP-independent document access; an archive system with its own logic for metadata administration and own authorization system is no longer required.
Next page: Allocation of tasks and the benefits of HSM
How tasks are allocated between SAP and archive
This equates to a paradigm shift in archiving: middleware is focused solely on the optimized administration of the technical metadata (archive date, storage period) and leaves the functional metadata such as customer name, account ID, invoice number, etc. to the SAP system. The middleware transfers the interfaces to the data and document sources (e.g. SAP), and the actual storage and media administration to the HSM system optimized for the purpose.
The benefits for the SAP user are seen in terms of more efficient usage of existing storage systems, less complex applications, and a reduced hardware overhead for the company, as leaner archives require fewer resources: an additional database to store the metadata is always required with traditional archive systems; this is no longer the case with the new approach. The access speeds are also improved since only the storage system has to be optimized to this end. Speeds with direct access to the HSM are well under a second per document – a figure that is almost always exceeded with traditional archive systems.
Why the traditional SAP user in particular benefits from HSM
Since legally compliant archiving of certain types of documents is now required by law, the new archiving type is in principle interesting for companies of any size, irrespective of the quantity of documents to be archived. The SAP penetration of the company is a much more decisive factor. Companies using just individual modules generally tend not to get involved with the issue of document archiving, even in the traditional way. The traditional SAP user by contrast utilizes their system universally and will have already established HSM infrastructures. For such a user, the new archive approach is therefore a compelling option. A ballpark figure of around 50 SAP workstations can be quoted based on real-life scenarios.