Yes. It worked..
Thanks a lot
regards
b s pradhan
---
On Wed, 03 Dec 2003 Tanel Poder wrote :
Try archive log stop; and archive log start; after setting the parameters..
Tanel.
Hi All,
When I set log_archive_dest_2 to a valid local destination
Dynamically on
Thanks to all who replied.
Alter table/index deallocate keep 64k worked like a charm.
I did not checked it in depth but I think that it keep 64k from the actual
end of the data, not the high water mark.
When I read the manual about keep nk I was sure that it means above the high
water mark.
Etagon invited me to come and visit them at their stand at the UKOUG
conference in Birmingham next week. Don't know if I'll have time or not,
but in general I'm still looking for hard evidence of financial savings
using RAC, ie a real comparison where switching to RAC (on whatever
platform)
Adam,
No, I don't think you can do it through DDL. For for regenerating and modifying the
code, the simplest to use is DBA_SOURCE.
HTH,
SF
- --- Original Message --- -
From: [EMAIL PROTECTED]
To: Multiple recipients of list ORACLE-L
[EMAIL PROTECTED]
Sent: Wed, 03 Dec 2003
It is not only database designers who have these problems - check out this
extraordinary verbatim quote from a certain well known international
politician...
Reports that say that something hasn't happened are always interesting to
me, because as we know, there are known knowns; there are
My understanding is that the parm simply reverts you
to pre-9 behaviour where index joins were not there,
so there should be no major problems.
hth
connor
--- April Wells [EMAIL PROTECTED] wrote:
We have discovered an... opportunity... and Oracle's
suggestion is to set
_index_join_enabled
Hi there ...
I have a table with a Varray Columns.
When I used a select operation without access the varray columns the
explain plan show fine.. cost = 4
but when I access the Varray columns then I got the cost = 1310
Why ? What happend ? and why oracle did it ?
Someone can explain it to me ?
He won a British award for that one ;) It beat the Schwarzenegger quote
I think that gay marriages should be between Men and Women ...
interesting point of view. (I live in California by the way :S)
-Original Message-
Robson, Peter
Sent: Thursday, December 04, 2003 2:39 AM
To:
Failed.
Raj
Rajendra dot Jamadagni at nospamespn dot
com All Views expressed in this email
are strictly personal. QOTD: Any clod
can have facts, having an opinion is an art !
-Original
Hi, friends:
Several months ago there is a thread talking about choosing the proper memory size
for windows server running oracle.
And today I logon to one of my small oracle on NT and found something I cannot
understand. It is a small application running Oracle 817/win2k.
SGA is
Here are all the details:
Source database 9.2.0.4 (upgrade from 8.1.6.3.4).
Target database 8.1.6.3.4.
View definition: create view my_view as select * from [EMAIL PROTECTED]
Sql: select * from local_table , my_view
where local_table.branch = 1
and my_view.customer = 200 +
Jonathan ... it is Microsoft product after all, you don't question why something
happens.
There is only one Golden Rule ... When in doubt, reboot ... and again
Raj
Rajendra dot Jamadagni at nospamespn dot com
All
I need a java package that will allow me to run OS commands(Unix) from a
stored procedure.
Anyone got one?
--
Please see the official ORACLE-L FAQ: http://www.orafaq.net
--
Author: John Dunn
INET: [EMAIL PROTECTED]
Fat City Network Services-- 858-538-5051 http://www.fatcity.com
San
John,
Try this:
Process p = Runtime.getRuntime().exec(/usr/bin/df -Pk);
Abraham
-Original Message-
Sent: Thursday, December 04, 2003 7:49 AM
To: Multiple recipients of list ORACLE-L
I need a java package that will allow me to run OS commands(Unix) from a
stored procedure.
Anyone got
expert one on one. check asktom.oracle.com might be on there also.
From: John Dunn [EMAIL PROTECTED]
Date: 2003/12/04 Thu AM 09:44:29 EST
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Subject: RE: java package to run OS command
Which book is that?
-Original
Hi List,
Does anyone use Change Manager software to keep track of the changes that went into
the database. We have 25 environments and having difficult time to keep track of the
changes between different release(version) within the same database and same release
across different database.
For
Hi,
But PGA is only 40M(This is the sum of all process's v$sesstat).
So there is more memory utilization then oracle actually should use.
From task manager, it is 2018(Physical+Virtual), But from oracle v$(sga + pga) it
is only 1020M.This is the problem.
Zhu Chao.
- Original
The parameter makes available an access method as Jonathan described, and it
*is* available in 8i, just defaults to false. And of course it isn't to be
confused with join indexes, the type of index spanning multiple tables,
introduced in 9i.
Here's an 8i example, note the VIEW OF
Cary, thanks a lot for the explanation (and the other
one [and especially the other one :-) ]). Much
appreciated.
--- Cary Millsap [EMAIL PROTECTED] wrote:
You're on the right trail by experimenting with it.
I'm surprised that
the number of gettimeofday() calls is so small. The
times() call
Hm...since I didn't run into this or any other problem installing
9.2.0.4 on Gentoo Linux plug plug, perhaps it's RHAS3-specific???
Rich
Rich Jesse System/Database Administrator
[EMAIL PROTECTED] Quad/Tech Inc, Sussex, WI USA
-Original
Solved.
It was a hash join with the smaller table first but it pulled the whole 1M
records for this.
There was a: where local_table.branch = 1 in the query. I changed it into:
where remote_table.branch = 1 (there is an index on remote_table.branch) and
it came down to 2 seconds.
Thanks all.
It's not the relink done during installation that's the problem, but running
a relink manually using the relink script in $OH/bin. I've seen this problem
on RHAS 2.1 and 3.0.
Sten
-Original Message-
Sent: Thursday, December 04, 2003 7:39 AM
To: Multiple recipients of list ORACLE-L
Yes. At least I was before reverting back to ANALYZE. For our little 30GB
DB, there's little penalty for non-parallel stats gathering on Saturdays.
Then again, one could ANALYZE multiple tables at once to regain some
parallelism in stats gathering...
Rich
Rich Jesse
I suppose just how much redundancy makes you sleep well is up to you. But,
one additional bit of info to keep in mind is that the backup info also gets
stored in the control files, and rman can use those too if no catalog
database is available. Making a copy of a control file after the backup
Title: RE: decrease initial extent
Mladen,
Keep reading, it's further down.
From the 8.1.7 SQL Reference Manual, under the ALTER TABLE entry
(top of page 8-36):
If you specify KEEP, then the specified amount of space is kept
and the remaining space is freed. When the remaining
number
Here's another link to a good Java Stored Procedure for this:
http://www.oracle-base.com/Articles/8i/ShellCommandsFromPLSQL.asp
But the same caveats apply. We're using an external procedure written in C for this
instead of Java, and I use a shell that checks PRODUCT_PROFILE for authority to run
cost is irrelevant. ignore it. it doesnt matter. its internal for oracle. what docs
are you using that say to use cost? none from oracle. they dont exist.
have your Logical I/Os gone up? Has your response time gone up?
I can guess as to why its more 'costly'? By accessing the varray do you do
Jared,
Thanks, but what I was wondering is how the views got created in the
first place. If the field name needs an alias (i.e. substr(x,i,j)) and you
include this alias in the CREATE VIEW ddl, it will also appear in
dba_views.text. If you don't include the alias in the DDL, the view won't
John, it is available on asktom as well, but read the caution Tom explains. Restrict
it to only executables you want. Asktom has example.
Raj
Rajendra dot Jamadagni at nospamespn dot com
All Views expressed in this
Title: oradebug suspend question
Well, we are running into a situation We apply code release to production environment once a month. During that timeframe, we still have db open and everyone connected and working. Sometimes what happens is sometimes code causes invalids and it takes
Which book is that?
-Original Message-
Sent: 04 December 2003 14:35
To: Multiple recipients of list ORACLE-L
its in tom kytes first book. might be on his webpage.
From: John Dunn [EMAIL PROTECTED]
Date: 2003/12/04 Thu AM 08:49:25 EST
To: Multiple recipients of list ORACLE-L
We have both 8i and 9i instances, but 'eventually' plan to migrate everything to 9i.
I'm looking at using RMAN for our backup and recovery. We have many instances but
essentially 2 types.
1. Production instances that have both OLAP and OLTP. These must be in archive log
mode.
2. We have
One I've used in the past was supplied by Tom Kyte :
http://asktom.oracle.com/pls/ask/f?p=4950:8:4723489521562622951::NO::F4950_P
8_DISPLAYID,F4950_P8_CRITERIA:952229840241,
Cheers,
Mike Hately
-Original Message-
Sent: 04 December 2003 13:49
To: Multiple recipients of list ORACLE-L
I
I do not think you need 2 Rman catalogs.
Build a small database just for Rman. After the daily backup finished bring
it down and do os level backup of all the environment for this small backup.
You can then easily restore it in case of disk problems.
Remember: A backup is good only when you can
i must have misread the docs. i thought it was either catalog or control file. didnt
know you could do both
thanks.
From: [EMAIL PROTECTED]
Date: 2003/12/04 Thu AM 11:04:26 EST
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Subject: RE: RMAN questions
I suppose just how
Henry--
Unless I misunderstand your question, the answer is that the aliases
are supplied in the CREATE VIEW statement _before_ the view's query,
e.g.
create view foobar (col1, col2) as select sum(bytes), sum(blocks) from
user_segments;
--- Poras, Henry R. [EMAIL PROTECTED] wrote:
Jared,
backup of all the environment for this small backup.
It should have been: for this small database.
Sorry about the mistake.
Yechiel Adar
Mehish
- Original Message -
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Sent: Thursday, December 04, 2003 6:44 PM
I do not think you
yeah, I'm trying to get away from C external procedures ...but java is
beginning to look just as much a hassle.
Why is running a OS command such hassle?
-Original Message-
Sent: 04 December 2003 15:40
To: Multiple recipients of list ORACLE-L
Here's another link to a good Java Stored
Raj, you are not suspending user sessions, you are suspending server processes
(ora_s000n_sid). That is not always safe, because other processes can also
send wake up signal to the shared servers. Unfortunately, oracle doesn't have
anything like suspend session, because it would be extremely
Here is a adoption for running the ls command ... be very careful with suspriv
command, it is very powerful privilege. One could replace ls with rm and then you are
doomed if oracle is running as root.
grant javasyspriv to system
/
create or replace type fileList as table of varchar2(255)
/
John,
I agree and am encouraging the external C procedures since their simpler, and
can create log files as well. BTW: Whoever authored the Java procedure on that web
page should be shot. I can just see someone passing 'rm -fr $ORACLE_HOME' to it.
Dick Goulet
Senior Oracle DBA
Oracle
MG,
I am not talking about alter system suspend I am talking about suspending USER
sessions
oradebug setmypid pid
oradebug suspend
...
oradebug resume
Raj
Rajendra dot Jamadagni at nospamespn dot
Hi.
The 2 GB process limit kicks in well under 2 * 1024 *1024 * 1024.
its between 1.7 and 1.8 GB.
I'm quite familiar with hitting it in win32, as large memory support was not enabled in every 8.1.7.x patchset. Large memory support sure works great in 9.2.0.4.
W2K3 Server (not Advanced) ships
Yechial,
It's been a couple of years since I worked on tuning queries with db links, but
a couple of issues come to mind:is the correct table being used for the inner
table of the join, is too much data being sent over the network.
-is the correct table being used for the inner table: I remember
Well, I'm going to get involved here saying upfront that my company is a
competitor of Etagon's, so I'm certainly biased, both about us vs. Etagon
and RAC in general.
However, the financial savings of RAC can be significant - we do cost
analyses all the time of RAC for potential customers, and
It really isn't that much of a hassle once you have Java itself setup.
I have a problem with the lack of security placed on the Java code from the
oracle-base.com article though. What we did (are in the process of doing)
is to restrict execute access to a single directory that we create and
Dick, harsh words, hmmm? Powerful tools can also be powerful weapons in the wrong
hands. But don't blame the toolmaker.
John, the reason that running an OS command is such a hassle is that it can be
horribly destructive to your server. An OS command that runs from a Java Stored
Procedure
will it catch following command apart from rm -rf ???
find /var/opt/oracle/logs -mtime +1 -type f -name *.trc|perl -nle unlink
Probably not ... and that's why it is dangerous ... basically you should have a set
of fixed programs that can be called and accept only arguments from calling
By default, no user accounts in the Oracle DB have any privs to do anything
to the OS. Privs -- including privs to execute, read, and write OS files --
must explicitly be granted (and rightly so!).
Do not grant the JAVASYSPRIV or JAVAUSERPRIV to the schema that will own the
ExecOS Java code, as
SGA is 970M and PGA(maxsize) is 40M. Connection is 20.But from task
manager, Oracle is using 1005M physical Memory and 1013M virtual memory(you
can view the data from here:
http://www.cnoug.org/html/ut/attach/2003/12/04/12516-oramem2-embed.gif).
Physical memory and virtual memory overlap in
In my opinion you're approaching the problem from wrong end. Something like
standby database, temporary BCV copy or similar sounds safe solution for me.
Or use mechanisms like deferring user reports during upgrades (or check for
some bit's value before running a report etc..)
Tanel.
MG,
I am
test
Mladen Gogala scribbled on the wall in glitter crayon:
That wouldn't do much harm since it's relatively static and probably
frequently backed up. Restoring $ORACLE_HOME would be an easy
excercise. Now passing datafiles or copying /etc/motd onto
/oradata/SID/system01.dbf would cause much more
Title: Message
Test failed
-Original
Message-From: Jeffrey Beckstrom [mailto:[EMAIL PROTECTED]
Sent: Thursday, December 04, 2003 1:55 PMTo: Multiple
recipients of list ORACLE-LSubject: testtest
No, but it disallows command chains - only single commands are permitted. If you want
to run more than one command you have to write it as a shell script, and the full path
to that shell script must be approved for the current user in the PRODUCT_PROFILE
table. Normally, we will only allow
Hi Justin
Didn't know you were on the list
A properly formed hint will cause the CBO to consider the
hinted path to be
less costly than it would otherwise consider it, but hints do
not force a
query to use that particular plan. For a moderately
complicated query,
you'd potentially
That is utterly disgusting memory management. When I come to think
of it, there was a guy named David Cutler who was promising that Windows
will have the same virtual memory system as VMS, with FREELIM,FREEGOAL,
BORROWLIM, GROWLIM and MPW_ parameters. Working sets are also gone as
well as the most
Title: Message
as I
understand it Oracle no longer uses (officially) processor speed to determine
license costs, though it appears that *actually* it uses revenue targets to
determine license costs
Niall
-Original Message-From:
[EMAIL PROTECTED] [mailto:[EMAIL
Can't do ... this is a 2 node RAC, users are connected all the time (basically this
app runs ESPN daily business). There is constant DML and reporting activity on the
database.
Raj
Rajendra dot Jamadagni at
Which wait events are indications that your missing bind variables?
btw, if you want to implement bind variables through a c/c++ middle tier its best to
use 'prepared statements'. correct?
bind arrays can be issued as prepared statements right? This is when you need to do
alot of inserts
http://www.dbazine.com/burleson18.shtml
Wow. We've got a lot of indexes to rebuild... :)
Seems like those criteria have been beaten up on this list before.
Rich
Rich Jesse System/Database Administrator
[EMAIL PROTECTED] Quad/Tech Inc, Sussex, WI USA
-Original Message-
Sent:
latch free waits on library cache shared pool latches.
But I'd check parse to execute ratio from v$librarycache instead. If most of
your statements close parse count and execution count, you might have a
problem.
Tanel.
- Original Message -
To: Multiple recipients of list ORACLE-L
Even though I have never touched VMS myself, I completely agree that it is
(was) a great operating system, I've just heard so many good words from
respectable sources about it :)
About Windows, probably the initial idea was great but since MS is a
marketing driven company, they just left off most
We're in complete agreement here. That, I believe was my suggestion, too.
On 12/04/2003 02:59:25 PM, Tanel Poder wrote:
In my opinion you're approaching the problem from wrong end. Something like
standby database, temporary BCV copy or similar sounds safe solution for me.
Or use mechanisms
There's no point of arguing about whether a query ran faster right after you
rebuilt your index. Nor there is no point in finding some ultimate algorithm
for finding the point of index rebuilding, we all know the answer - it's
42.
Instead, a long stress test has to be done, e.g. running 10
Hey all,
I'm testing out UltraSearch with 9.2.0.4 (non-9iAS) and the default Apache
that comes with it on Gentoo Linux 1.4 with a do_brk-patchless
2.4.20-gentoo-r8 kernel. So far, it ain't bad at all. However, I do have
some Adobe PDFs that come across without links when they're listed as a
Btw, the one good use for plan stability stored outlines are poorly
written 3rd party applications, where you can't even add a hint to code. I
these cases you can compose a good execution plan yourself and set your
statement to use it.
Tanel.
- Original Message -
To: Multiple
Then have two sets of packages for reports, when releasing new versions, do
them on set 2 whilst normal users are running on set 1. This gives you time
to test new versions in production as well. When tests are successful,
switch to set 2.
Oh yes, this requires your application/reports to be
Speaking of Apache, does anyone have a good makefile to
link MOD_PLSQL into Apache? Does anybody know how to get
mod_plsql into Apache2?
On 12/04/2003 04:09:28 PM, Jesse, Rich wrote:
Hey all,
I'm testing out UltraSearch with 9.2.0.4 (non-9iAS) and the default Apache
that comes with it on
Thanks,
Added this one to my bookmarks along with
http://www.oracle.com/oramag/webcolumns/2001/index.html?auto_index.html
and
http://www.dba-oracle.com/art_index1.htm
Sten ;)
-Original Message-
[EMAIL PROTECTED]
Sent: Thursday, December 04, 2003 11:55 AM
To: Multiple recipients of
I know I've posted this before, but it's been many years, so here we go
again.
NT was supposed to be Windows' answer to VMS. WNT, doesn't stand for
anything, so how did they come up with the name?
V+1=W
M+1=N
S+1=T
Just like
I-1=H
B-1=A
M-1=L
Coincidence?
Bambi.
-Original Message-
Hi List,
I have OEM which is running on my local NT box. OEM repository is located in my local
database/NT box.
I am connecting to all the databases which is running on Solaris thru tunneling for
all DML and DDl activity.
Now I would like to add the databases which are running on Solaris to my
Comments in-line
Regards
Jonathan Lewis
http://www.jlcomp.demon.co.uk
The educated person is not the person
who can answer the questions, but the
person who can question the answers -- T. Schick Jr
One-day tutorials:
http://www.jlcomp.demon.co.uk/tutorial.html
Three-day seminar:
we plan to implement ORACLE data guard under 9iR2. We
have experience on 8i standby database, but never
touch 9i data guard. Does anyone have document how to
implement 9ir2 standby databse under Data Guard?
Thanks.
__
Do you Yahoo!?
Free Pop-Up Blocker - Get it
This:
http://otn.oracle.com/pls/db92/db92.show_toc?partno=a96653remark=drilld
ownword=dataguard
Is probably a good place to start.
Mark J. Bobak
Oracle DBA
ProQuest Company
Ann Arbor, MI
Imagination was given to man to compensate him for what he is not, and
a sense of humor was provided to
Hi ??
who's asking? dba1 mcc is rather anonymous :-(
I've implemented over 20 DG installations as far, based on
AIX/Solaris/Linux/W2K. The first Business Critical Zero Dataloss
installation went live in November 2001.
I haven't used the Oracle GUI for dataguard, simply because that didn't
While working on some scripts to monitor PGA usage on 9i, I came across
something interesting while experimenting with different forms of SQL.
I have recently been forcing myself to make use of 'OVER..PARTITION BY' in
SQL so as to be more comfortable in using it. Can't add new tools to the box
While playing around with SQL for some PGA scripts, I managed to create some SQL
that will consistently cause ORA-600 [kkqwrm_noref: COLFDNF set]
This appears to be Bug # 2507421, which was supposedly fixed in 9.2.0.3.
Here's the SQL:
select
pga_target_for_estimate
,
You are right Jared.
ORA-00600: internal error code, arguments: [kkqwrm_noref: COLFDNF set],
[], [], [], [], [],[], []
this is on oracle 9.2.0.3 Win2k+SP3.
Regards,
Jp.
-Original Message-
While playing around with SQL for some PGA scripts, I managed to create
some SQL
that will
Rhojel - I'm not sure if export gives you enough information. Usually the
reason to export is that if you can export, you have a good copy of the data
and you can breathe a little easier. But usually export will fail on the
first corrupt block. Your corrupt block may be an index block, so export
1. latch free for the shared pool latch (a shared pool latch in 9i).
2. Indirectly, SQL*Net message from client and CPU service time
consumption too, because not using bind variables probably means that an
application parses more often than it needs to.
Cary Millsap
Hotsos Enterprises, Ltd.
alert/trace files ?
- Original Message -
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Sent: Thursday, December 04, 2003 9:34 PM
Anyone know what might be causing this error? Oracle 8.1.7 on Solaris.
background process LGWR did not start
--
Please see the
Hi Gurus-
Peformence of my DB(8174, Tru64 Unix) has terribly slowed down!
Checked Indexes, Stats, etc but all look fine to me.
Only one strange thing:
I have more than 100 sessions in the v$session wait with same EVENT,P1TEXT,
P1,STATE
(SQL*Net message from client ,driver id,1413697536,WAITING).
Thanks Dennis! I've read a couple of docs on getting enough information on the corruption.
I'd like to use DBMS_REPAIR.CHECK_OBJECT, but I'd like to know first if this would be safe
to run on an online production database. Could anyone comment on this?
Oh Dennis, sorry for the nightmares ;-)
Hi Ganesh-
Thanks for your highlight.
From thei query,
select s.sid SID,s.serial# Serial, s.osuser O/S-User, s.username
Ora-User,
s.process Process-ID, s.status
Status,sw.state,sw.seconds_in_wait,sw.event,s.program,s.machine,s.terminal
from v$session s,v$session_wait sw
where sw.sid=s.sid
order by
86 matches
Mail list logo