Something l would try in that position is
create stored outline on ...
your select statement in 8.1.6
Then select from user_outline_hints
to find what hints Oracle had generated
in the fast plan. Then write the same
set of hints into the SQL in 9.2 (or even
export/import the hints) - someti
PROTECTED]]
Sent: Tuesday, March 11, 2003 12:00PM
To: Multiple recipients of list ORACLE-L
Subject: Re: Oracle position on hints
Seems like we are starting YetAnother Bind Discussion!
Anyway, the peeking into bind variables isbasically done when a hard
par
I doubt there are "some incorrectly applications", I think that is a rule
rather than an exception.
Raj
-
Rajendra dot Jamadagni at espn dot
com Any views expressed here are
strictly personal. QOTD: Any clod can
have facts, havin
.
Jared
"Jesse, Rich" <[EMAIL PROTECTED]>
03/10/2003 01:39 PM
To: "'[EMAIL PROTECTED]'" <[EMAIL PROTECTED]>
cc: "'[EMAIL PROTECTED]'" <[EMAIL PROTECTED]>
Subject:RE: Oracle position on hi
After spending 2 ½ days trying hints, init parameters, re-writing the query, a
completely useless TAR, etc.
to get a query that runs in < 1 second on 8.1.6.X to go faster than 1 minute on
9.2.0.2
I found a new to 9.2.0.X dynamic init parameter " optimizer_dynamic_sampling", if I
understand it
Title: RE: Oracle position on hints
Hi,
We recently upgraded from 7.3.4 to 8.7 (management plans on getting to that Y2K problem shortly ;-) We had an SQL statement that really needed a hint in 7.3.4. After upgrading to 8.7, I removed the hint and it runs much faster without it. I spend
Sent by: cc:
[EMAIL PROTECTED] Subject: RE: Oracle position on hints
Sent by: [EMAIL PROTECT
Star transformations still seem to be disabled
by bind variables in predicates against the
dimension tables, even in 9.2. (I have an idea
that materialized views have the same problem.
but I haven't tested that).
Also, bind variable peeking isn't much help
with histograms as the peek only takes
03 01:39 PM
To: "'[EMAIL PROTECTED]'" <[EMAIL PROTECTED]>
cc: "'[EMAIL PROTECTED]'" <[EMAIL PROTECTED]>
Subject:RE: Oracle position on hints
I thought that went out the window with 9i because it can snoop at t
stograms.
Jared
"Jesse, Rich" <[EMAIL PROTECTED]>
03/10/2003 01:39 PM
To: "'[EMAIL PROTECTED]'" <[EMAIL PROTECTED]>
cc: "'[EMAIL PROTECTED]'" <[EMAIL PROTECTED]>
Subject:RE: Oracle p
cc:
Subject:RE: Oracle position on hints
Given the attitude of most dba's that you should always use bind variables
where possible I can't see how you'd cope with skewed data without them.
Most developers should know a databases's data better than the o
Congratulations !
I had never thought that my question on hints would have generated so much
trafic.
Stephane
-Original Message-
McDonald
Sent: Saturday, March 08, 2003 8:54 PM
To: Multiple recipients of list ORACLE-L
Just to add my meaningless diatribe to the discussion,
I think that
Oh, I don't know. There is the FIRST_WIVES(1) hint which should always be employed.
There is also the rules-based SWAMBOO (She Who Always Must BE Obeyed Optimization)
mode. I'm sure others as well.
Congratulations Connor!
Ian MacGregor
-Original Message-
Sent: Saturday, March 08, 2
;t be used with bind variables.
Jared
"Nicoll, Iain" <[EMAIL PROTECTED]>
Sent by: [EMAIL PROTECTED]
03/08/2003 06:23 AM
Please respond to ORACLE-L
To: Multiple recipients of list ORACLE-L <[EMAIL PROTECTED]>
cc:
Subject:RE:
Congrats to you from us in the Freeman household too, Connor!!!
On another topic, bought your book on Perl Jared We shall see
if you can make a Pearler out of me yet...
RF
-Original Message-
To: Multiple recipients of list ORACLE-L
Sent: 3/9/2003 11:53 PM
Congratulations Connor!
W
I think it must have. He certainly hasn't been on email since!;)
Pete
"Controlling developers is like herding cats."
Kevin Loney, Oracle DBA Handbook
"Oh no, it's not. It's much harder than that!"
Bruce Pihlamae, long-term Oracle DBA
-Original Message-
Hamilton
Sent: Monday, Marc
Congratulations Connor. May the day exceed all your expectations. If you
figure out how to attach a piece of cake to an email, send me one. :-)
--
Chuck
- Original Message -
To: "Multiple recipients of list ORACLE-L" <[EMAIL PROTECTED]>
Sent: Saturday, March 08, 2003 8:53 PM
> Just to ad
Congratulations Connor!
We shant tell your bride how you whiled away the
hours on your wedding day. ;)
Jared
On Saturday 08 March 2003 17:53, Connor McDonald wrote:
> Just to add my meaningless diatribe to the discussion,
> I think that there are "good" hints and "bad" hints.
>
> For example,
Title: RE: Oracle position on hints -
Congratulations Connor !!
Raj
This e-mail
message is confidential, intended only for the named recipient(s) above and may
contain information that is privileged, attorney work
Congratulations!
No hints for you:)
Regards,
Waleed
-Original Message-
To: Multiple recipients of list ORACLE-L
Sent: 3/8/03 8:53 PM
Just to add my meaningless diatribe to the discussion,
I think that there are "good" hints and "bad" hints.
For example, first_rows(n) is a good hi
Congratulations Connor!!! This is a great day in your life, although some guys would
not agree...
And thanks for sharing your knowledge about Oracle.
-Original Message-
Sent: Saturday, March 08, 2003 8:54 PM
To: Multiple recipients of list ORACLE-L
Just to add my meaningless diatribe
Connor, what can we say? You have gotta getta life!
Well, since you're getting a wife instead, that should take care of that
anyway. Best wishes to you and Gillian, hope the day is all you want it
to be.
Pete
"Controlling developers is like herding cats."
Kevin Loney, Oracle DBA Handbook
"Oh no
Just to add my meaningless diatribe to the discussion,
I think that there are "good" hints and "bad" hints.
For example, first_rows(n) is a good hint, in that you
are giving the optimizer more information as to what
your motivations are for this query, without stifling
it into a choice that may no
It's like the Auto-pilot
-Original Message-
To: Multiple recipients of list ORACLE-L
Sent: 3/8/03 2:13 PM
Stephane
I don't see how Oracle could ever have an official position on hints.
If
they say hints are going away, many developers panic. If they say hints
are
great, then Oracle co
Stephane
I don't see how Oracle could ever have an official position on hints. If
they say hints are going away, many developers panic. If they say hints are
great, then Oracle competitors will claim Oracle has a bad optimizer.
A key political issue in some shops is "Using hints makes our co
Given the attitude of most dba's that you should always use bind variables
where possible I can't see how you'd cope with skewed data without them.
Most developers should know a databases's data better than the optimiser and
certainly when building queries it's always worthwhile seeing where data i
Title: RE: Oracle position on hints
The hints,
actually, but then I haven't heard anything of the instructors going away
either! J
Pete
"Controlling developers is like herding cats."
Kevin Loney, Oracle DBA Handbook
"Oh no it's not. It's
I have heard from Oracle that you should stop using RULE hints as the
Rule-based optimizer (RBO) will eventually be retired or totally
unsupported.
We are moving away from analyze as well and using dbms_stats instead because
analyze is also going away. Migrating to dbms_stats is a first step for
There's always the possibility of a gap between
what was said and what was heard.
In my seminar, I tell people about hints, and
mention a useful ones - and then say they should
avoid using them if possible. Perhaps the quoted
instructor said something more like "you shouldn't
use hints in 9i if
Hi,
I have to write a SQL and PL/SQL guide for our developpers.
Some are good but most of them can use help.
I'm not a big fan of hints except for insert append, I used them only if I
really need to.
In last october I went to a 3 days DB2 UDB course, unless I miss something
there are no hint wit
Title: RE: Oracle position on hints
Pete
when you say ... "I haven't heard anything about them going away." you mean the hints or them instructors ??
Raj
-
Rajendra dot Jamadagni at espn dot com
Any views exp
Title: RE: Oracle position on hints
I heard this from a COTS vendor once. I also thought it was ridiculous. It is one of the useful ways to tune SQL Statements. I have only gotten the impression from their inception into the Oracle RDBMS that they were advocated by Oracle.
-Original
You just gotta wonder about some instructors. :(
I haven't heard anything about them going away.
Pete
"Controlling developers is like herding cats."
Kevin Loney, Oracle DBA Handbook
"Oh no it's not. It's much harder than that!"
Bruce Pihlamae, long-term Oracle DBA.
-Original Message
have facts, having an opinion is an art !!
-Original Message-From: Khedr, Waleed
[mailto:[EMAIL PROTECTED]Sent: Friday, March 07, 2003 12:25
PMTo: Multiple recipients of list ORACLE-LSubject: RE:
Oracle position on hints
The first day
hints are gone, I will change my
The first day
hints are gone, I will change my career!
I do not think it
will go away, since Oracle itself uses them to control the way PQ slaves and
other background processes work.
Also stored
outlines is part of it.
Regards,
Waleed
-Original Message-From: Stephane Paqu
Ive seen Tom Kyte post on his site that he doesnt like hints. However, even in 9i
there are times when I need to use hints to improve performance.
Hi,
Does Oracle have an
official position on hints ?
Will they go away as
the optimiser is becoming bettre or they are there to stay ?
TIA
This is facinating, where does this come from? This is the second time in a
week that I've heard this statement being made by someone. Hints will never
go away IMHO, and Oracle continues to add more and more of them. I was told
by someone at a client site that they were told by an Oracle instructor
38 matches
Mail list logo
