relational round robin archive

Hi,
could someone please point me in the direction of resources describing
the design of a round robin archive in a relational database? It
obviously is not common practice but still it should be possible. I
woud be after a draft for a schema and mechanisms which need to be
implemented for the aggregation of data to different granularities the
older data becomes. My objective is to copy some of the properties of
the RRA which is e.g. the limitation in growth and fixed max size of
the data being collected.
many thanks,
Christoph.