[comp.databases] Performance of views in relational database

wongj@hpsgm2.sgp.hp.com (Jimmy Wong) (10/04/90)

I am designing a relational database that need to join information from 
quite a number of tables. Can someone tell me will there be a difference in
performance if I join the information using a view as oppose to joining the
information at run-time ? Thank you in advance.


--
Jimmy Wong                                  phone: +65 279-8755
Singapore Networks Operation            Unix mail: wongj@hpsgm2.sgp.hp.com
Hewlett-Packard                           
1150 Deport Road
Singapore 0410

davidm@uunet.UU.NET (David S. Masterson) (10/05/90)

In article <1070003@hpsgm2.sgp.hp.com> wongj@hpsgm2.sgp.hp.com (Jimmy Wong)
writes:

   I am designing a relational database that need to join information from 
   quite a number of tables. Can someone tell me will there be a difference in
   performance if I join the information using a view as oppose to joining the
   information at run-time ?

In theory, there will be little difference as the information in the view must
still be processed by the query optimizer.  What can help the joins more is
clustering that some relational database systems implement.  That is, some
database systems allow keys from multiple tables to be clustered in the same
data pages which reduces the number of disk accesses because it will be more
likely that, after fetching A to join to B, B will be in memory.  Your mileage
may vary from system to system.

--
====================================================================
David Masterson					Consilium, Inc.
uunet!cimshop!davidm				Mtn. View, CA  94043
====================================================================
"If someone thinks they know what I said, then I didn't say it!"