Friend :
You can usethe option INDEXFILE of imp
program.
INDEXFILE=c:\your_name.txt
Ity will generate the script.
I am not sure if it runs with exp, may be
...
I had used with imp.
Regards
Eriovaldo
- Original Message -
From:
Santosh
Varma
To: Multiple
Someone else cracked it -- he doesn't have TIMED_STATISTICS = TRUE...
- Original Message -
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Sent: Saturday, October 26, 2002 9:53 PM
*no* waits?
How is this possible?
Is intantaneous computing now a reality?
Sorry for the
DENNIS WILLIAMS wrote:
Stephane
You mentioned each additional index costs about 2.5 times the cost of
inserting into a non-indexed table. I just wanted to point out that Kevin
Loney has done some performance tests involving the number of indexes. I
don't know if he has published these
Umm,
The program sets up an ODatabase object and then references
ODynaset objects to issue queries
e.g. assuming you have a 3 edit control login dialog box with a
class name of MYDialog. Somewhere in the InitInstance() member
for the application object you should have
#include oracl.h
#include
Naveen -
He provided figures, and they are on the handout that is somewhere in my
office. Maybe I'll run across it someday, or even better, maybe he'll
publish his results. Here are a few more details that I posted to this list
earlier.
Kevin Loney (author of Oracle DBA Handbook) has
PiotR
Welcome to the list. I don't work with Linux, but it shouldn't be
different from any other Unix in this aspect.
First, make sure your environment variables for ORACLE_HOME and ORACLE_SID
are set correctly. Your PATH variable should include $ORACLE_HOME/bin.
From the command line,
But even with TIMED_STATISTICS=FALSE, the system will still have waits.
Just because the duration of said waits is not accurately recorded
doesn't mean they do not exist. And if one looks at the V$ wait
interface, the waits will be logged there.
-Mark
On Sun, 2002-10-27 at 09:48, Tim Gorman
Of course the system will still have waits -- it's just that the query
provided will show values of zero for all of them when TIMED_STATISTICS =
FALSE. That's why Rahul said that there were no waits at all. My bad...
Rahul,
Please enable the parameter TIMED_STATISTICS = TRUE, re-run the INSERT
I believe Kevin gave that presentation at OpenWorld -- either last year
or the year before. His paper is available for download on the TUSC
site, as he works for TUSC.
--- DENNIS WILLIAMS [EMAIL PROTECTED] wrote:
Naveen -
He provided figures, and they are on the handout that is somewhere
The way I solved this problem is to create a useable dummy DB_LINK so
that compiling the code will succeed.
The only minor gotcha is that what will result if/when a bogus dblink is
passed into the code for real at run time.
HTH YMMV
HAND!
I've been trying off on for about 5 MONTHS to get 9iAS R2 into a useable
condition.
Either I'm suffering from a bad case of senioritis or this s/w should be
classified as User Hostile s/w NOT ready for prime time.
I'd like to hear from anybody who has this POS installed in production;
Just an idea but how about using Java Stored Procedures? They would allow
better control over OS files. Also they would be housed within the
database. Combine this with some Oracle XML technology your work could be
made alittle easier. Although memory intensive.
From: Jared Still [EMAIL
You can read all about it at:
http::/www.oreilly.com/catalog/oracleperl
and
http://www.amazon.com/exec/obidos/tg/detail/-/0596002106
:)
Jared
On Saturday 26 October 2002 20:59, Leonard, George wrote:
Hi Jared
Someone mentioned that you have a perl+dba tool set, would you mind telling
us
Hi All,
I want to feed my shell script with the names of currently running
databases. I thought of using ps -ef|grep [p]mon. What I got was the
following:
oracle 20113 1 0 Oct 25 ?0:01 ora_pmon_TLDEV
oracle 898 1 0 Jul 22 ?0:06 ora_pmon_TLQA
oracle 944
When you tried under NT, did you use sqlplus.exe or sqlplusw.exe?
If you used sqlplusw.exe, this sounds like Bug No. 1504702
The workaround is to use the copy command from the character version of sqlplus
(sqlplus.exe) rather than the GUI version.
HTH,
Bruce Reardon
-Original Message-
As is often the case, there are a million ways to do this. Given the
output listed, I'd do something like:
ps -ef|grep pmon|grep -v grep|awk '{ print $8 }'|awk -F_ '{ print $3 }'
ps -ef -- get the programs that are running
grep pmon -- get only those lines that have 'pmon' in them
grep -v grep
;-)
the only waits i see are parallel query dequeue wait, and sometimes
v$session_wait shows write complete waits ..
but i'm sure these are not slowing down the process... (or are they ?)
as most of the time v$session_wait does not return a row !!
and i query this view once every second
top waits from system_event, when i siad no waits .. i ment no waits while
querying session_wait !!
offcourse a DB WILL experience waits... but, are these waits slowing down my
inserts ??
these stats are after the insertion of 16 million rows, the table in
question is the only table on that disk
Hi Ross,
Try something like this:
DBNAME=`ps -ef | grep pmon | awk '{db=substr($NF, 10); print db}'`
There are several ways of doing this - what's above is just one.
Cheers
Sujatha
-Original Message-
Sent: Monday, 28 October 2002 12:43 PM
To: Multiple recipients of list ORACLE-L
Hi
Thanks Mark.
Yes that helps, in a way. I could use good ol' awk to parse the last field!
However, a slight problem in the ps -ef output. If I do what you have
suggested on my 'ps -ef' output below, I would only get RMAN as it is the
only one in field #8. All the rest are in field #9. All I
ps -ef|grep [p]mon | awk '{ print $7 }' | sed 's/ora_pmon_//'
On Sun, 27 Oct 2002, Ross Collado wrote:
Hi All,
I want to feed my shell script with the names of currently running
databases. I thought of using ps -ef|grep [p]mon. What I got was the
following:
oracle 20113 1 0 Oct
Browsing through my Sed and Awk book following up on Mark's lead, I came up
with this:
ps -ef|grep pmon|grep -v grep|awk '{ print $NF }'|awk -F_ '{ print $3 }'
TLDEV
TLQA
TLQAVAR
DBMON
RMAN
PRDINF
TLDVVAR
That'll do fine !
Thanks again.
Ross
-Original Message-
From: Ross Collado
Yes this works too. I think the trick is in deriving the last field in this
case it was the $NF all along.
Thanks to all who replied and those who attempted to.
Ross
-Original Message-
From: Sujatha Madan [mailto:Sujatha.Madan;CMC.OPTUS.NET.AU]
Sent: Monday, 28 October 2002 14:38
Hmm...good point. You know, I use this technique in various places. I
never noticed that bug before, cause it will only appear if the database
has been up less than 1 day. Guess it's time to do a bit of script
auditing.
Ok, try this instead:
ps -ef|grep pmon|grep -v grep|cut -c49- |awk -F_ '{
24 matches
Mail list logo
