54 PART II Prerequisites for an HCM
✔✔The selection of data to be archived. Not all data in an HCM system needs
to be archived. For instance, it is sufficient to archive a document only in its
original format if it meets the requirements for long-term archiving. The
selection must be rule-based and fully automated in the background. This
requires a set of rules that can be flexibly adapted and can use all relevant
document metadata as criteria for selection.
✔✔Determination of the retention period. An HCM system stores data for
which different legal retention periods apply (e.g. certificate of incapacity
for work for one year, X-ray for ten years or for children up to 28 years of
age). The HCM system should automatically determine the retention period
based on rules.
✔✔Deletion of data. The principle of storage restriction (formerly data economy)
is reaffirmed in the EU data protection basic regulation. After the
statutory retention periods have expired, there is often no further reason to
keep the data available: it must be deleted. The HCM system must therefore
delete all formats of this data, including the archived data. With WORM
storage systems that support the concept of retention time, the HCM system
can initiate the deletion of this data at the end of the retention period. If
retention periods are not supported by your WORM storage, we advise you
to distribute the data to storage media when archiving so that you can make
entire storage media inaccessible after the retention period has expired.
✔✔Creation of backups. If archiving is the sole use case for data, the data to be
archived can be moved to long-term storage. If you want to maintain
high-performance access, a backup copy for the long-term storage and a
copy for fast access are recommended.
✔✔Ensuring archiving and restoring data. An HCM system must ensure that
the data actually ends up on the long-term storage. It goes without saying,
but it is often underestimated: it must be possible to recover data efficiently
from long-term storage.
✔✔Data modification. Experience has shown that data that has already been
archived can still be modified at a later time. An HCM system should therefore
be able to propagate modifications to the long-term storage or apply
them while restoring data.
✔✔Consistency verification. In an ideal archive, the data is completely secure.
But there are no ideal archives, neither digital nor paper-based. In fact, errors
creep in over time for a variety of reasons. In digital archives, data consistency
can be checked with reasonable effort, and errors can be easily