Avnish,
The book goes through this a bit on p61. There are a few tools out
there, including three from Oracle. As I mention in the book, I use our
own Hotsos Profiler, described at www.hotsos.com/products/profiler.html.
Cary Millsap
Hotsos Enterprises, Ltd.
http://www.hotsos.com
Upcoming events
On Fri, 12 Dec 2003 [EMAIL PROTECTED] wrote:
> I will try to get the output of v$system_event and will send it you guys. In the
> mean time I have more question..
>
> I am reading Cary's 'Optimizing Oracle Performance Book'. I am half way thru and
> over looked rest of the chapters but didnt fi
I will try to get the output of v$system_event and will send it you guys. In the mean
time I have more question..
I am reading Cary's 'Optimizing Oracle Performance Book'. I am half way thru and over
looked rest of the chapters but didnt find an easy way to analyze thousands of lines
trace file
All,
This sounds wy too familiar to me.
My (blind) guess is that sql*net round trips is killing performance.
System-wide could indicate this, but, as Jared states, trace out a specific session, and grab the session-specific info from v$sesstat, before and after.
We brute forced the issue of
The wholesale system wide collection of timing data is not generally
a good way to go about trouble shooting performance issues.
You need to pick a process, collect the timing data for that process,
and *only* that process, diagnose where the most time is being spent,
and determine what can be don
**
>-Original Message-
>From: [EMAIL PROTECTED]
>[mailto:[EMAIL PROTECTED]
>Sent: Thursday, December 11, 2003 11:29 AM
>To: Multiple recipients of list ORACLE-L
>Subject: RE: Performance tuning in complex environment
>
>
>Not really sure what happened
Avnish - Since nobody has mentioned it yet (my posts arrive late, so
probably will by the time this appears), get Cary Millsap's book Optimizing
Oracle Performance
http://search.barnesandnoble.com/textbooks/booksearch/isbnInquiry.asp?userid
=6WIANMIL0H&isbn=059600527X&TXT=Y&itm=1
His methods soun
:www.compuware.com
-Original Message-
[EMAIL PROTECTED]
Sent: Thursday, December 11, 2003 2:29 PM
To: Multiple recipients of list ORACLE-L
Subject:RE: Performance tuning in complex environment
Not really sure what happened and why we decided to that. I was involved in
Oh I've run into THIS beforeyou are in a sticky technical AND political
situation I am sure.
It is really not that complex.
I'll bet they (your DBAs) have already been told you that the app is
horribly designed and it was a mistake and that the hardware is under
powered canned "dataserver ins
Thanks
I asked because we also use Citrix and so far we never had a problem related to
Citrix, only problems we had were inefficient coding and oracle bugs, nothing related
to HW/disk/WTS etc. The only problem initially with Citrix was configuring client
printers, but our guys figured it out a
Not really sure what happened and why we decided to that. I was involved in the
beginning of project and remembered that PM was mentioning about talking to another
Logician client who were facing same issues.
-Original Message-
Jamadagni, Rajendra
Sent: Thursday, December 11, 2003 10:55
DBAs should never 'guess' about performance. If they are guessing you need new DBAs.
They should be running statspacks, sql trace, and looking at timing data.
Its too much to explain in an email. Fire your DBAs and find people who dont 'guess'.
How much are you paying these guys?
>
> From: <
Ummm ... what was the problem that prompted you guys to replace citrix servers?
Raj
Rajendra dot Jamadagni at nospamespn dot com
All Views expressed in this email are strictly personal.
QOTD: Any clod can have facts,
Cary,
I detoured from the new Tom Kyte book after chapter 4 to read your test through.
chapter 6 of the tom kyte book might have been a better band aid for me at the moment, but - I finally (4 weeks after the issue was raised) got a user to let me know when he was going to run a posting routine.
Ryan,
Your two questions have different answers.
I studied mathematics as an undergrad. I focused on the abstract stuff:
predicate calculus, language theory, functional analysis, topology,
In my studies I constructed many, Many, MANY proofs. (A "proof" in
mathematics is a piece of technical
what is your math background? what level of math would you recommend
performance specialists to have?
- Original Message -
To: "Multiple recipients of list ORACLE-L" <[EMAIL PROTECTED]>
Sent: Tuesday, October 21, 2003 3:49 PM
> Michael, I've responded by preceding your questions with "MM:
here is a list of tuning books to read. I used to work with the guy who
wrote it. He definitely knows what he is doing. There are quite a few people
on this list who can attest to that.
http://www.amazon.com/exec/obidos/tg/listmania/list-browse/-/VL8CI2YJANX1/re
f=cm_lm_dp_l_2/102-3468524-1000163
Cary,
Thank you for your in-depth response. It was very helpful. To me, the
hardest books to read and understand are those that tell you WHAT but not
WHY. From the excellent reviews I've received (look at MLaden's review just
posted), it appears to give plenty of WHY. I appreciate that very much.
Thank you Dennis. I will take a look at that sample chapter, then probably
go out and pick up the book.
Thanks again.
This e-mail, including attachments, may include confidential and/or
proprietary information, and may be used only by the person or entity to
which it is addressed. If the reader
Cary, I believe that I'm more then entitled to a commission.
On 10/21/2003 06:04:26 PM, Michael Milligan wrote:
Cary,
Thank you for your in-depth response. It was very helpful. To me, the
hardest books to read and understand are those that tell you WHAT but
not
WHY. From the excellent reviews I've
Sorry to double post. It didn't show up on the board and after about an hour
I thought there was a problem. Of course as soon as I posted again, they
both showed up! I'll be more patient next time.
Michael Milligan
Oracle DBA
Ingenix, Inc.
2525 Lake Park Blvd.
Salt Lake City, Utah 84120
wrk 801-98
Cary,
Thank you for your in-depth response. It was very helpful. To me, the
hardest books to read and understand are those that tell you WHAT but not
WHY. From the excellent reviews I've received (look at MLaden's review just
posted), it appears to give plenty of WHY. I appreciate that very much.
t; From: "Cary Millsap" <[EMAIL PROTECTED]>
> Date: 2003/10/21 Tue PM 03:49:24 EDT
> To: Multiple recipients of list ORACLE-L <[EMAIL PROTECTED]>
> Subject: RE: Performance tuning book
>
> Michael, I've responded by preceding your questions with "MM:&qu
I was fortunate to attend the Hotsos Clinic and read the advance version. It is
not your normal tuning book. If you are looking for checklists, hand holding,
"do x to solve y" solutions, don't bother with the book. If you are looking for
a book to teach you the skills to solve your performance prob
ap" <[EMAIL PROTECTED]>
> Date: 2003/10/21 Tue PM 03:49:24 EDT
> To: Multiple recipients of list ORACLE-L <[EMAIL PROTECTED]>
> Subject: RE: Performance tuning book
>
> Michael, I've responded by preceding your questions with "MM:" and my
> answers with &
Michael, I've responded by preceding your questions with "MM:" and my
answers with "CVM:".
MM: ...can you please tell me if your new book, of which I've heard good
things, is different in any way than other Oracle Performance Tuning
books out. Does it take a different approach?
CVM: Drastically
Thank you Dennis. I will take a look at that sample chapter, then probably
go out and pick up the book.
Thanks again.
This e-mail, including attachments, may include confidential and/or
proprietary information, and may be used only by the person or entity to
which it is addressed. If the reader
Sorry to double post. It didn't show up on the board and after about an hour
I thought there was a problem. Of course as soon as I posted again, they
both showed up! I'll be more patient next time.
Michael Milligan
Oracle DBA
Ingenix, Inc.
2525 Lake Park Blvd.
Salt Lake City, Utah 84120
wrk 801-98
Michael
Oh yeah, this book takes a revolutionary approach compared to any other
book written to this point. The first chapter is posted at
http://www.oreilly.com/catalog/optoraclep/index.html
read this and you will see that this book is an entirely new method of
locating the root cause of Oracle
Trace the slower procedure. Hit the 10046 paper on www.hotsos.com to see
how.
This sounds like maybe 'buffer busy wait' waits on the index are causing
contention among the procedures. But you need to prove whether it is
(and which block it is, if my guess is right) before you can take the
right co
This is where the balancing comes into picture. If it is a bulk Insert then
definitely the performance would degrade to the extent that it has to create
an entry in the Index at a particular place. If there are so many indexes on
this table you should visualise them in such a manner that a concaten
Perhaps you should consider disabling / dropping the index during the first
procedure and then recreating it. You can use commands like 'execute
immediate' within a procedure to issue DDL.
Otherwise perhaps you can change the inserts to some kind of bulk insert -
depending on how your application
Vivek - Hopefully you will receive some replies from someone with RAC
experience. However, since RAC is so new, the information on tuning it may
be pretty slim. Since RAC is based on Oracle Parallel Server, you might
consider searching for tips on OPS. Some tips might apply to RAC.
Dennis Williams
Sorry for the late reply,
thx a lot all of u for the help
Regards
Sameer
--- "Godlewski, Melissa"
<[EMAIL PROTECTED]> wrote:
> Sam,
>
> Que1:
> If you have statistics on the table then the cost
> based optimizer is used.
> If you have out of data statistics you could be
> sending bad information
Sam
Que 3 - I have no idea how the parameters you are passing affect the
execution time. That one probably needs more explanation.
Que 5 - You don't say what you are performing backups with. RMAN? System?
Hot? I am going to assume that you are doing hot backups (archivelogging)
with the O.S. (no
Here are replies to a couple Sam,
1) If optimiser mode is chose and any table in a query has stats then CBO
will be used. If stats do not exist on any table Oracle has a hard-coded
default value which is 100 if I recall a note on here correctly.
If stats do exist but they are very old then they w
well, we would certainly expect *you* to know this :)
thanks, that helps when I send links to my mom, who still feels she has
to retype the url rather than cut and paste
--- Jeremiah Wilton <[EMAIL PROTECTED]> wrote:
> A note about Amazon URLs:
>
> You can often cut off a lot of the stuff at t
A note about Amazon URLs:
You can often cut off a lot of the stuff at the end when sharing a
link with others. The stuff at the end is just session context. For
instance, Chris's list is at:
http://www.amazon.com/exec/obidos/tg/listmania/list-browse/-/VL8CI2YJANX1
--
Jeremiah Wilton
http://ww
Commit; :-)
And then go to OraPerf.com, Hotsos.com, Jonathan Lewis' website with one
of these wonderful, long and strange English names, Ixora, etc. and get the
good articles there. This is where you'll find a lot of current thinking.
Mogens
[EMAIL PROTECTED] wrote:
Okay, I'll plug my ow
Okay, I'll plug my own list of recommendations on Amazon. It
should come up from this URL:
http://www.amazon.com/exec/obidos/tg/listmania/list-browse/-
/VL8CI2YJANX1/qid=1013704150/sr=5-1/ref=sr_5_1/103-8633316-
6595843
If it doesn't, just enter 'Oracle performance tuning' in a search on
Amazon
RKAR, Samir" <[EMAIL PROTECTED]>
Sent by: [EMAIL PROTECTED]
01/31/02 03:20 AM
Please respond to ORACLE-L
To: Multiple recipients of list ORACLE-L <[EMAIL PROTECTED]>
cc:
Subject:RE: Performance Tuning
Okay Rachelu have me convinced there
[EMAIL PROTECTED] wrote:
>I agree we need to look at both. Depends on the shop you work in if you
>can though. Where I've been the SAs hold tuning of the OS close, and I
>can work WITH them but can't effect changes on my own. so I concentrate
>on the database.
>
it also helps if the OS is tunable
I agree we need to look at both. Depends on the shop you work in if you
can though. Where I've been the SAs hold tuning of the OS close, and I
can work WITH them but can't effect changes on my own. so I concentrate
on the database.
Besides, I've always found it much more effective if I go to the
Okay Rachelu have me convinced thereI am adding the 101 book to
my shelf !!
I said *some* of the masters of OracleI do realize that there r lots
of other masters of Oracle who do not work for TUSC.
Yep...Alomari's book does deal more on OS tuning for Oracle but don't u
think
that
Rafiq,
actually I was in a good mood... I really DO want specific instances of
things that you think need more explanation or something we didn't
include that you'd want to see.
But "some things should be discussed in detail" doesn't tell me WHAT
things and I can't fix anything.
Rachel
--- Moh
Rachel,
It looks that you are not in good mood today. I got a message from
Kirti,also,on same subject and I am replying to him directly
Regards
Rafiq
Reply-To: [EMAIL PROTECTED]
To: Multiple recipients of list ORACLE-L <[EMAIL PROTECTED]>
Date: Wed, 30 Jan 2002 11:37:25 -0800
rather tha
rather than being coy and saying "some aspect/topics are discussed in
detail" why not say what you'd like to see?
First -- Oracle Press strictly limits the number of pages in their
books, it is a constant challenge to cram all the information that the
authors want into the page limit.
Second --
Hi Rafiq,
You are most welcome to write to us directly in detail your
comments/suggestions.
Many have done that.
We will make every effort to incorporate our readers' suggestions when we
work on the 2nd edition.
Thanks.
Best Regards,
- Kirti
-Original Message-
Sent: Wednesday, Janua
I didn't say Rich's book is bad, just that if you could only buy one,
I'd buy the 101 book. It's written in a more "friendly" style than the
other one, mostly because that's the goal of the 101 series.
Of course, since I did some of the editing on the book, I'm prejudiced
(and no, I don't get any
the soul of diplomacy, that's me :)
I am prejudiced, I did edit the book after all. But I don't endorse
what I don't believe in. Everyone I have talked to within and without
Oracle is espousing the tuning by wait events methodology that drives
your book.
and thank you for proving the rollback/o
Very cute ---your submisness
This book may be improved further if some aspect/topics are discussed in
detail..
Regards
Rafiq
No... we are not.. still at #1.
Must be something wrong with K.Gopal's keyboard or mouse (or finger;)
I sincerely thank you all for 'discussing' our book.
Gaja is still
eshpande, Kirti" <[EMAIL PROTECTED]>
>Reply-To: [EMAIL PROTECTED]
>To: Multiple recipients of list ORACLE-L <[EMAIL PROTECTED]>
>Subject: RE: Performance Tuning
>Date: Wed, 30 Jan 2002 09:10:33 -0800
>
>Rachel,
>Thank you so much for you kind words and
No... we are not.. still at #1.
Must be something wrong with K.Gopal's keyboard or mouse (or finger;)
I sincerely thank you all for 'discussing' our book.
Gaja is still catching up with his office e-mail after returning from a
long vacation.
But he will join the list soon..
Regards,
-
Rachel IS kind but her words are also true and I think this was her
main focus in this case. ;-)
My Alomari book was for v7 and was good but is now dated...
Is it up to date for 8i/9i?
-Original Message-
Sent: Wednesday, January 30, 2002 10:11 AM
To: Multiple recipients of list ORACL
[EMAIL PROTECTED]]
> Sent: Wednesday, January 30, 2002 6:30 AM
> To: Multiple recipients of list ORACLE-L
> Subject: RE: Performance Tuning
>
> Oracle Performance Tuning 101- Oracle Press - Gaja, Kirti
> Oracle Performance Tuning 101- Oracle Press - Gaja, Kirti
> Oracle
Rachel,
Thank you so much for you kind words and the very interesting comparison..
:)
- Kirti
-Original Message-
Sent: Wednesday, January 30, 2002 9:50 AM
To: Multiple recipients of list ORACLE-L
I would perhaps suggest Oracle Performance Tuning 101 instead of Rich's
book (if you had t
Maybe u r right, Rachel.since except for me, all the others have
suggested the 101 book. Thing is, I bought Rich's book based on the
customer reviews on Amazon and I must say that I am not disappointed but
then,
I haven't read the 101 book. At the same time, TUSC is supposed to have some
of
I would perhaps suggest Oracle Performance Tuning 101 instead of Rich's
book (if you had to make a choice).
the 101 book conforms most closely to Oracle's current thinking on
performance tuning
--- "SARKAR, Samir" <[EMAIL PROTECTED]> wrote:
> Practically, u should get two books to have a full
Yeah its very good book!!!Especially I like that chapter Method behind
madness!!..Every body should read that book.
>From: "K Gopalakrishnan" <[EMAIL PROTECTED]>
>Reply-To: [EMAIL PROTECTED]
>To: Multiple recipients of list ORACLE-L <[EMAIL PROTECTED]>
>Sub
Hi,
are they up to release 5 already?
I'm 4 releases behind :-))
Jack
"K Gopalakrishnan" <[EMAIL PROTECTED]>@fatcity.com on 30-01-2002 13:30:23
Please respond to [EMAIL PROTECTED]
Sent by: [EMAIL PROTECTED]
To: Multiple recipients of list ORACLE-L <[EMAIL PROTECTED]>
cc:(bcc: J
Oracle Performance Tuning 101- Oracle Press - Gaja, Kirti
Oracle Performance Tuning 101- Oracle Press - Gaja, Kirti
Oracle Performance Tuning 101- Oracle Press - Gaja, Kirti
Oracle Performance Tuning 101- Oracle Press - Gaja, Kirti
Oracle Performance Tuning 101- Oracle Press - Gaja, Kirti
Best
Binay,
The Oracle Performance Tuning book from O'Reilly (ISBN 1-56592-237-9) is an
excellent book. Covers just about everything, database, code, environment
tuning.
Regards,
Kev.
-Original Message-
Sent: 30 January 2002 10:25
To: Multiple recipients of list ORACLE-L
Hi Everyone
Practically, u should get two books to have a full grasp on Oracle
Performance
Tuning :
1. Oracle Performance Tuning Tips and Techniques by Richard Niemiec
2. Oracle 8i and Unix Performance Tuning by Ahmed Alomari
If u r in an exceptionally high development environment where u have lots of
SQL
www.oracletuning.com has a few interesting scripts, Kym.
-Original Message-
Sent: Monday, July 23, 2001 2:06 PM
To: Multiple recipients of list ORACLE-L
Does anyone have any good ideas on where to start
tuning and any good scripts out there.
thanks
kym
www.ixora.com.au
www.vampired.net
O'Reilly Performance Tuning
O'Reilly Oracle Internals
Oracle 8i and Unix Performance Tuning
"Walking on water and developing software from a specification are easy if
both are frozen."
Christopher R. Spence
Oracle DBA
Fuelspot
-Original Message-
Se
"Oracle Performance Tuning 101" from Oracle Press. Scripts are in the
book. Hang out a while and you will see posts to this list from the
authors...
Scott Shafer
San Antonio, TX
k johnson wrote:
>
> Does anyone have any good ideas on where to start
> tuning and any good scripts out there.
>
ORACLE-L <[EMAIL PROTECTED]>
<[EMAIL PROTECTED]cc:
mpo.gc.ca> Subject: RE: Performance tuning
sorry, I should have been clear, here is the quote from the article...
"RAID 5 costs more for write-intensive applications than RAID 1. The
so-called "small-write penalty" inherent in the design of RAID level 5 disk
arrays
makes each Oracle DBWR write require four physical I/O operations.
Conse-q
ilto:[EMAIL PROTECTED]]
|| Sent: Wednesday, May 30, 2001 4:21 PM
|| To: Multiple recipients of list ORACLE-L
|| Subject: RE: Performance tuning
||
||
|| Translation please.
||
|| Alex Hillman
||
|| -Original Message-
|| Sent: Wednesday, May 30, 2001 3:32 PM
|| To: Multiple recipients of
t is
simply interesting and tangentially connected to
metaphor generation and usage.
hth,
Ross
|| -Original Message-
|| From: Hillman, Alex [mailto:[EMAIL PROTECTED]]
|| Sent: Wednesday, May 30, 2001 4:21 PM
|| To: Multiple recipients of list ORACLE-L
|| Subject: RE: Performance
[mailto:[EMAIL PROTECTED]]
|| Sent: Wednesday, May 30, 2001 1:22 PM
|| To: Multiple recipients of list ORACLE-L
|| Subject: RE: Performance tuning
||
||
|| Why putting DB in noarchivelog will avoid redo generation?
|| There will be no
|| archiving, but redo will continue to be generated
You have to get up pretty early in
the morning to pull the wool over
your eyes, Mr. Hillman.
:)
|| -Original Message-
|| From: Hillman, Alex [mailto:[EMAIL PROTECTED]]
|| Sent: Wednesday, May 30, 2001 1:22 PM
|| To: Multiple recipients of list ORACLE-L
|| Subject: RE: Performance
Yes, Agreed.
Raid of any level will not change the number of writes in terms of DBWR.
But will change how many physical reads and writes needed to implement
parity and data to the disks.
"Walking on water and developing software from a specification are easy if
both are frozen."
Christopher R.
I don't know what is causing this, but I would keep an eye on physical
memory available vs. commit charge in Task Manager, I strongly recommend
your commit charge never exceed your physical memory.
I found that Oracle sometimes refuses to even start services when it runs
out of physical memory, i
No to be overly pedantic, but RAID5 does not change the number of write()
calls made by DBWR. It will change the number of disk operations done by
your hardware controller or your software raid drivers, but that's slightly
different.
- Original Message -
To: "Multiple recipients of list
have no disk limit
problem. i have autoexetended datafiles.
Regards
Azhar
Peter McLarty
incts.com> cc:
Sent by: Subject: Re: Pe
Sorry - that's what I meant. No archive log generation. NOT no redo log
generation. Thanks for pointing that out. Is it Friday yet?
Jim
> Why putting DB in noarchivelog will avoid redo generation? There will be
no
> archiving, but redo will continue to be generated.
>
> Alex Hillman
>
>
Indeses are not critical for the application to run..It just helps them run faster at
times. You could build you indexes with the "no logging" or "unrecoverable" option
and that will save on the archivelog creation. Just be sure to take a backup after the
indexes are rebuilt because you will n
Redo generation will not change. Just the archiving process, which on
improperly laid out file system can be very painful process.
"Walking on water and developing software from a specification are easy if
both are frozen."
Christopher R. Spence
Oracle DBA
Fuelspot
-Original Message---
Raid 5 can consume as much as 60% overhead for writes.
"Walking on water and developing software from a specification are easy if
both are frozen."
Christopher R. Spence
Oracle DBA
Fuelspot
-Original Message-
Sent: Wednesday, May 30, 2001 12:51 PM
To: Multiple recipients of list ORAC
John, have you read Cary Millsap's paper on RAID 5? www.hotsos.com I
believe. RAID 5 will require additional writes by DBWR which is going to be
huge overhead during a dataload.
- Ethan Post
-Original Message-
Sent: Wednesday, May 30, 2001 6:51 AM
To: Multiple recipients of list ORACLE-
Why putting DB in noarchivelog will avoid redo generation? There will be no
archiving, but redo will continue to be generated.
Alex Hillman
-Original Message-
Sent: Wednesday, May 30, 2001 10:21 AM
To: Multiple recipients of list ORACLE-L
Azhar,
Just two thoughts off the top of my hea
Azhar,
Just two thoughts off the top of my head to improve loading of records:
1. Put database in noarchivelog mode to avoid redo generation (if
possible).
2. Drop all indexes on target tables (if possible), then load data, then
recreate all indexes on target tables.
Jim
Jim Hawkins
Lead SA
t you need to get some other system stats (i/o, cpu, waits, locks) and
configuration data (SMP?, RAID??).
Rodd Holman
>>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<
On 5/30/01, 6:21
incts.com> cc:
Sent by: Subject: Re: Performance tun
I agree. Is your Shared_pool_size set to 239MB? You can't ask Oracle to
allocate this much memory with excessive swapping on your NT box. If you
can't add mode RAM, I would cut that to around 50 MB. I would also do what
Peter recommends below. There are always a few more things to try (like
direct
FOR YOUR INFORMATION
ESIS and EPFAL are now part of Logica. The Internet email addresses of the staff has
changed to the following - [EMAIL PROTECTED] eg [EMAIL PROTECTED] Emails
using the old format will continue to be delivered until 30th June 2001.
Peter McLarty wrote
"Are you running RAID
I would not recommend using optimal as it will cause the segment to have to
shrink. Your better of doing it manually on lulls of activity and size the
segments bigger if need be. Segments should not be frequently shrinking.
"Walking on water and developing software from a specification are ea
Azhar,
I agree with the other responses (more RAM, don't use OPTIMAL on your
RBS's) but would also suggest dropping all indexes except PK until the load
is done, the rebuild them. Also, does this table have any
triggers? If so, if can you disable them during the load, that will help
spe
Okay fill us in why would you try and inane thing like that on only 128MB
RAM. NT would be barely running on that without getting your instance up. I
am not that good yet at calculating what amount of RAM you need for a given
SGA but I am thinking that you are running in swap
What is the disk
90 matches
Mail list logo
