"George McLean" <gmclean (AT) rrsytec (DOT) com> wrote

before the point in time the update occurred, and then roll the logs forward
from the backup to just before the point in time when the update occurred.
To do this, it is necessary to previously have log retain enabled.

Figuring out the point in time to recover to is a bit more problematic. In
theory there is a entry in the log for every update, but it may not be
possible to figure out how to read the log to identify the transaction point
in time you need. The data in the log is designed to be used by DB2 internal
processes and not users or administrators. But if there is only one update
for the entire object (table) in question during a known timeframe, you
might be able to find it. Finding when a specific row was changed is more
difficult. Check out manuals for more information about reading the log
data.