At 02/10/2002 10:05 AM (Sunday), mmenaz wrote:

>... but a
>huge number of unindexed reading to RDB$RELATION_CONSTRAINTS (588 for a 10
>rows
>result dataset).
>What does it mean? How I can have better performance on
>RDB$RELATION_CONSTRAINTS reading? (I mean, can I add indexes? Indexing what?)

I suspect that the row count for this table is relatively low (hundreds) so
the whole table is most likely in memory cache all of the time. Therefore,
the fact that it is scanned many times may not be a real performance
problem. It more likely indicates simply that the data in the table is used
many times in your stored procedure and trigger logic.

Are you seeing a performance problem that you think points to this table?
Or, are you just looking at the access count and wondering what the need
for it is?