In the event that the DQguru finds an unusually high number of matches
(Jonathan said that this would be roughly over half the number of toal
records or higher), performance gets degraded significantly, largely
because waiting to to transfer the result table data over the network. It
also slows down viewing results in the match validation mode, and also
loading the match pool in subsequent match engine runs.
In such situations, it's probable that the user's transformations are
either too loose, resulting in too many matches, or their data has serious
quality issues. In any case, it would be worth notifying the user that
their transformation is finding an unusually high number of matches, and
prompt them to either continue storing these matches into the result
table, or to stop the engine entirely so that the user can tweak their
transformation.
Original issue reported on code.google.com by mo.j...@gmail.com on 9 Sep 2009 at 7:25
Original issue reported on code.google.com by
mo.j...@gmail.com
on 9 Sep 2009 at 7:25