--
"Teppicamon" <quicog_at_hotmail.com> wrote in message
news:bdpnrv$v0qdd$1_at_ID-162849.news.dfncis.de...
>
> "Burton Peltier" <burttemp1REMOVE_THIS_at_bellsouth.net> wrote in message
> news:XR9La.3227$kx.673_at_fe04.atl2.webusenet.com...
> > --
> > "Noons" <wizofoz2k_at_yahoo.com.au.nospam> wrote in message
> > news:3efc225d$0$8262$afc38c87_at_news.optusnet.com.au...
> > > "Teppicamon" <quicog_at_hotmail.com> wrote in message
> > > news:bdh5r1$srdu7$1_at_ID-162849.news.dfncis.de...
> > >
> > > > Not at all, believe me... We wrote one application using SQLServer
> that
> > made
> > > > some quite complex queries on a materialized view resulting of
joining
> > five
> >
> > Wouldn't you have to use MVs for queries in sqlserver because otherwise
> you
> > would have locking problems on the underlying tables ?
>
> Not at all, we just used MV because they gave us better performance and
let
> us index everything exactly the way we wanted. That's completely opposite
to
> the MV's in Oracle, which at least in 8i turned to be even slower than
doing
> the joins in real time (maybe it was just that we didn't do it the right
> way, but we had to go back to normal joins). That's actually the point
with
> SQL Server, it's easy, it's clear, it's powerful enough and everything
> behaves as expected...
>
> >
> > > > tables, one of them of about 50 million rows (guess the size of the
> > view,
> > > > then ;-) ). Results come to the web application users (with the
> network
> > and
> > > > asp interpretation overhead) in less than two seconds...
> > >
> > > I'd hate to think how long it would take Oracle if
> > > Billy had counted these using a MV... ;)
> >
> > Probably takes Oracle about 50 msecs (milliseconds for you sqlserver
> users),
> > especially if a sqlserver MV is the same as an Oracle MV ?
>
> From my experience, SQLS MV's behave far much better than Oracle's, and
for
> your words I wonder if you're just guessing about SQL Server without
having
Good guess. I don't have much experience with sqlserver. That is why I
stated the 'IF' in front of 'a sqlserver MV is the same as an Oracle MV'...
because I do not know.
From all the posts since mine, seems you needed to tune sql queries against
the Oracle MV and would probably have had similar performance.
> any experience with it... Why everything has to be faster in Oracle just
for
> being Oracle??? You sound like a blind Oracle talliban shouting "Oracle
has
> to be the best because Allah told us one century ago... Death to the
> unbelieving!!!" ;-)
Where did I state anything at all like this? What the locking comment?
Everyone knows this is a problem in sqlserver and not a problem in Oracle.
Nothing new there. What did I say... to be associated with the talliban ?
Note: Show me a better product and I would gladly switch to it. I have
switched OSs, databases, and programming languages so many times, I could
not guess the number of times. But, switch from Oracle to sqlserver does not
make any sense at least in my work environment (and a lot of others I'd
guess).
>
> > The comment about 1 of 5 tables being 50 million rows means nothing. The
> MV
> > could be 100 rows after doing the joins and then it's just a 100 row
table
> > (in Oracle anyway) ? So, then 2 seconds would be realllly slow , IMHO.
>
> Blah, blah, blah... When I said it was a join, I meant it was a join. And
> when I let to your imagination the total number of rows, I guessed
everyone
> would understand that the number would be several times 50 million rows...
> Not the case as I see. But just for the sake of information, I'll tell you
> it was over 400 millions rows... Now go and make a complex query against
> that MV and tell me if 2 seconds (and in these 2 seconds you have to
include
> network and asp overhead) is slow...
>
> All the best
>
>
>
>
Received on Tue Jul 01 2003 - 20:51:00 CDT
