On Sun, Oct 17, 2004 at 09:06:56PM -0400, Paul Appleby wrote:
I have a simple Perl 5.6 test script that uses DBI and DBD::Oracle to connect to a
local Oracle 9i database table and retrieve the data in the three small fields of
its only two records.
Why is the connection time so long and
Paul Appleby wrote:
I have a simple Perl 5.6 test script that uses DBI and DBD::Oracle to
connect to a local Oracle 9i database
[...]
Why is the connection time so long and how can I shorten it?
[...]
It takes 2.9342188835144 seconds to connect to the database.
How long does it take with
Well, do you know what cause that delay? Between an Oracle client and
Database Server?
- Original Message -
From: Steffen Goeldner [EMAIL PROTECTED]
To: Paul Appleby [EMAIL PROTECTED]
Cc: [EMAIL PROTECTED]
Sent: Monday, October 18, 2004 11:34 AM
Subject: Re: Slow connection to Oracle 9i
file: $CPAN/authors/id/S/SG/SGOELDNER/DBD-ADO-2.93.tar.gz
size: 42606 bytes
md5: c3cad1e4b5576895939423f8130da1bb
Changes:
Implemented do().
Changed $sth-DESTROY to call finish( $sth ) instead of $sth-finish.
The old call caused clearing of errors e.g. in $dbh-do().
Replaced
Hi
I am using Oraperl emulation for DBD:Oracle for a perl program in HP-UX11i.
I would like to know the amount of CPU usage by Oraperl since we are having
CPU usage constraints.
I would also like to know whether waiting for input will put the state of the
process to
Hi
I am using Oraperl emulation for DBD:Oracle for a perl program in HP-UX11i.
I would like to know the amount of CPU usage by Oraperl since we are having
CPU usage constraints.
I would also like to know whether waiting for input will put the state of the
process to
On Mon, Oct 18, 2004 at 02:20:14PM +0530, rajarathnam, devaraj wrote:
Hi
I am using Oraperl emulation for DBD:Oracle for a perl program in HP-UX11i.
I would like to know the amount of CPU usage by Oraperl since we are having
CPU usage constraints.
The CPU usage by Oraperl (plus
Tim,
It is 3 to 4 times longer than retrieving data from a MySQL database.
Only 3 to 4 times longer than retrieving data from a MySQL?
You're lucky, it's often longer! :)
That was 3 to 4 times longer using the data from d:DProf.
But it's 20 times longer using Time::HiRes to measure the time it
On Mon, Oct 18, 2004 at 10:38:23AM -0400, Paul Appleby wrote:
DBD::Oracle::dr::load_dbnames is only called by data_sources()
so don't call data_sources() unless you really need to.
I really do need to call data_sources() but the time it takes to
retrieve data, as shown above, using
I'd like to get a list of open source perl (DBI based) database
interface GUIs for unix (Tk/Gtk etc).
If people could send me (off-list) details of any they know
I'll summarize back to the list.
Thank you.
Tim.
Success at last! I had the dba create an instance with the database
character set AL32UTF8 (thanks to the excellent documentation with
DBD::Oracle 1.16, which explained the weird behavior of Oracle with plain
UTF8), and used NLS_LANG=.UTF8, and DBD::Oracle 1.16 (1.15 did not work
100%, even
Hi !
I have this message chen I tape the line :
ppm install DBI.ppd
Error : no suitable installation target
What is the reason of this message. I didn't find anything on the web.
Could you help me please ?
Thanks, Raph
-
Créez
Most people with experience with Oracle know that opening oracle connections
is SLOW!
Oracle does not appear to consider that a problem, just like they do not
consider slow performance for doing DDL a problem
Applications that require near real time (OLTP) response times open
connections once,
Steffan,
Try measuring the time it takes to connect with SqlPlus by using
Time::HiRes with the variables $time1 and $time2 both set to
gettimeofday() and the the length of time set to $time2-$time1, as
follows:
$time1=gettimeofday();
your code runs here;
$time2=gettimeofday();
print
Maybe Oracle does not think it is a problem, but I (and lots of others) would
disagree. These issues are correctable.
In the case of slow logins, it depends on what is being done when the connection is
made. When making connections that are dedicated server (not MTS) a process is
created.
On Mon, Oct 18, 2004 at 11:48:11AM -0600, Reidy, Ron wrote:
1. Read the DBD::Oracle docs. There is a section (albeit, dated) that describes
how one might make connection times faster.
2. Read the Net Services Admin Guide
On Mon, Oct 18, 2004 at 05:09:05PM +0100, Tim Bunce wrote:
I'd like to get a list of open source perl (DBI based) database
interface GUIs for unix (Tk/Gtk etc).
I should clarify that a little. Basically I'm intereted in something
that'll open a dialog box asking for connection details, then
My CGI application will be called by different users at different
times. Are you saying the first user's connection can be left open
for all the other users? How?
Paul
Most people with experience with Oracle know that opening oracle connections
is SLOW!
Oracle does not appear to consider that a
Perhaps this'll help:
http://perl.apache.org/docs/1.0/guide/performance.html#Persistent_DB_Connections
There's a link on that site to Tim Bunce's Advanced DBI talk, but it returns a 500.
Tim???
Rich
Rich JesseSystem/Database Administrator
[EMAIL PROTECTED]
No. Each user will be using the same application that logs in to the
same schema. But the users are all anonymous visitors to a web site.
Paul
Are you logging in each user to a unique oracle schemna?
If so, no hope... (other than oracle tuning per Tim's message --
pre-spawned listeners on the
Look at using Apache::DBI for persistent connections.
-
Ron Reidy
Lead DBA
Array BioPharma, Inc.
-Original Message-
From: Paul Appleby [mailto:[EMAIL PROTECTED]
Sent: Monday, October 18, 2004 2:39 PM
To: BAXTER, LINCOLN A
Cc: [EMAIL PROTECTED]
Subject: RE: Slow
H...one thing I thought of is to check the validity of the handle, in case the
Oracle instance bounces (i.e. The Oracle instance is available, but the persistent
connection no longer exists). Does that automagically happen in the connect or should
there be code to check for a specific
I have a DBIx::HA (High Availability) module on CPAN, but I'm readying
a next version real soon now that will take care of this for you.
If you combine Apache+mod_perl+Apache::DBI+DBIx::HA, you'll get what
you need.
I'm debugging one last instance of a dbh having an ActiveKid handle
when it
On Mon, 18 Oct 2004 16:51:35 +0200 (CEST), Raphael wrote:
Hi Raphael
ppm install DBI.ppd
Error : no suitable installation target
It's been years since I had this problem, so I'm really guessing, but I think it's due
to a conflict between the Perl you are using and the Perl used to prepare
On Mon, 18 Oct 2004 19:24:03 +0100, Tim Bunce wrote:
Hi Tim
Can't be
web-based.
Oh :-((.
--
Cheers
Ron Savage, [EMAIL PROTECTED] on 19/10/2004
http://savage.net.au/index.html
I am using DBI and DBD-ODBC to connect to an MSSQL Server and this query
gets an invalid cursor state:
PRINT 'starting select'
select count(*) from anytable
PRINT 'finished'
If I only run the PRINT command... it works.
If I run the full query against a sybase server... only the top select
On Mon, Oct 18, 2004 at 04:39:25PM -0400, Paul Appleby wrote:
My CGI application will be called by different users at different
times. Are you saying the first user's connection can be left open
for all the other users? How?
Apart from the other (good) advice here, which you should follow
No reauthenticate() method in DBD:Oracle v 1.12 or in the ActiveState
Perl 5.6 docs.
Is there a Windows IIS equivalent for the Apache::DBI for persistent
connections?
Is there a way to configure or set Oracle form its Enterprise
Management Console to establish a persistent connection to a
I realize I may not have been totally clear in my earlier post, which
should have said:
My CGI application will be called by different visitors to the web
site at different times, but they all connect to the same database
using the same connection variables, i.e. the same user/password.
Are
Paul,
Without knowing much about Oracle (although I can talk your ear off about Sybase), you
want to cache connections to the server and set a limit (maybe its 1!) and use this
pool of connections to connect to your server.
Major advantage:
No overhead for creating the connection on
my $sql = qq#
PRINT 'starting select'
select count(*) from anytable
PRINT 'finished'
#;
is not valid sql.
try this:
my $sql = qq#
select count(*) from anytable
#;
my $sth = $dbh-prepare( $sql );
print 'starting select';
my $rv = $sth-execute();
while ( my $hr = $sth-fetchrow_hashref ){
Jeff,
It is VERY VALID syntax for both MSSQL ... AND.. Sybase. I have been coding
in Perl and DBI for several years now and I am NOT trying to use the perl
print function. Here is the skinny on PRINT direct from our DBA for MSSQL
Server PRINT :
The PRINT statement takes either one character
James
if you think it's a DBD::ODBC issue, why not use DBD::Sybase instead?
Dan
Moosmann, James [EMAIL PROTECTED]
19/10/2004 16:17
To: '[EMAIL PROTECTED]' [EMAIL PROTECTED]
cc: '[EMAIL PROTECTED]' [EMAIL PROTECTED], (bcc: Dan
Horne/IT/AKLWHG/WHNZ)
Dan,
The server and all the SP's I am running this on is MSSQL and not Sybase. I
have another DSN to a Sybase server and tested this behavior on that one,
and reported the results here. I was hoping to find someone who knew the
ODBC driver guts well enough to know what the problem was or knew a
Please read the documentation for Apache::DBI.
In Apache 1.3 and mod_perl 1.xx, when using Apache::DBI in conjunction
with DBI you get the following:
Each Apache process runs one perl instance. Each perl instance loads up
one instance of Apache::DBI. Each instance of Apache::DBI automatically
35 matches
Mail list logo
