Lately we've experienced an increase in data rows from 15 million to 35
million daily.

The processing time has gone from 20 minutes to nearly three hours.
Aggregates have remained the same.

Am I experiencing data explosion?

you have to reevaluate your aggregations because the old ones are too old.
the aggregation design is no longer optimal for the sizing.

also you have to play with your memory parameters because the new volume
required a different set of memory parameters.

you can also consider partitionning your cube. (3 * 15millions should be 3 *
20 minutes...)


"Jesse O." <jesperzz (AT) hotmail (DOT) com> wrote

we partition by day, and there's no really other option....35 million rows a
day now.

i redesigned the aggs, went form 65 down to 20....not much of a difference
however.

which memory parameters?


"Jeje" <willgart (AT) hotmail (DOT) com> wrote