![Wink :wink:](./images/smilies/icon_wink.gif)
Dave's original question is about removing 1/65th of the data weekly and adding 1/65th more. Could you justify a reorg for 1/65th of your data being deleted, if you consider database downtime during the reorg as well as the nodes being slammed for the duration? If you reorg every 4th week, at least you've moved to 1/16th and can schedule it around users needs.
My preference is to never delete as part of the ETL, but to update with a logical delete (a delete indicator column) so that the operation is reversible should you later discover errant data, or simply wish to audit the process. Then, during a maintenance cycle physically delete the older logically deleted data and then if the delete/non-deleted ratio reaches a threshold then conduct the reorg.
The strategies definetely have to be worked out. Great discussion!