Re: Why this query is sooo slow?!

From: Romeo Olympia <rolympia_at_hotmail.com>
Date: 21 May 2004 23:43:16 -0700
Message-ID: <42fc55dc.0405212243.47d11244@posting.google.com>

Things we know, we think we know, and we could try knowing:

(1) The large number of rows involved (particularly tables HIGH and GEOIPADDRESS_M) might be too much for the NL operations being done. Hence, the slow execution.

(2) Hash joins cannot be used by CBO because the join between those 2 large tables is not an equijoin.

(3) Can you try to test if you could hint a merge join instead? /*+ use_merge (...) */. Note that we just wanna test this and see if it yields better performance. We'd rather not use hints in production really.

(4) It's the weekend and I need to run some errands. ;o)

Cheers,

Romeo

ibm_97_at_yahoo.com (JZ) wrote in message news:<10bc841c.0405210712.642f826a_at_posting.google.com>...
> Still the same with this index. Thanks.
>
> > Its your query with all your info applied.
> > I wonder if ip.block is indexed.
> > Try to create index on that column.
Received on Sat May 22 2004 - 01:43:16 CDT