On Tuesday 04 April 2006 01:37, John Russell wrote:
> sub authenticate {
>
> my $r = shift;
>
> my $dbh;
>
> $dbh = DBI->connect("dbi:Oracle:XE", "WEB_LOGIN", "",
> {RaiseError => 0, AutoCommit => 1, PrintError => 0})
> || return "Error connecting.\n$DBI::e
Original Message
From: John Russell <[EMAIL PROTECTED]>
Cc: modperl@perl.apache.org
Subject: Re:pnotes MP1->MP2
Date: Mon Apr 03 2006 20:08:13
Here's the whole entry in the apache error_log
[Mon Apr 03 05:00:35 2006] [error] access to
/home/httpd/htdocs/com.onlywebdata.mc3/
Tom Williams wrote:
> Hal Vaughan wrote:
>> I'm using Apache 2.0 on Debian Linux. I've installed mod_perl and I can
>> see
>> the links for perl.load and perl.conf in /etc/apache2/mods_enabled. My
>> perl.load is like this:
>>
>> LoadModule perl_module /usr/lib/apache2/modules/mod_perl.so
>>
>>
Hal Vaughan wrote:
> I'm using Apache 2.0 on Debian Linux. I've installed mod_perl and I can see
> the links for perl.load and perl.conf in /etc/apache2/mods_enabled. My
> perl.load is like this:
>
> LoadModule perl_module /usr/lib/apache2/modules/mod_perl.so
>
>
>SetHandler perl-script
>
I'm using Apache 2.0 on Debian Linux. I've installed mod_perl and I can see
the links for perl.load and perl.conf in /etc/apache2/mods_enabled. My
perl.load is like this:
LoadModule perl_module /usr/lib/apache2/modules/mod_perl.so
SetHandler perl-script
PerlResponseHandler ModPerl::Regi
On Apr 3, 2006, at 8:19 PM, [EMAIL PROTECTED] wrote:
I don't think there's any down-side to adding DISTINCT. (Is it
standard
sql? I think it probably is, but I don't have a reference.) If there
is, it would be worth adding another yes/no knob to the config.
SELECT DISTINCT behaves radica
[EMAIL PROTECTED] wrote:
I just noticed one additional tweak that would be worth adding to your list:
It would be good to add "DISTINCT" to the group select statement... That
is, change (after my patch):
my $select= "SELECT $Attr->{grp_field}, $Attr->{uid_field}";
to:
my $select= "S
I just noticed one additional tweak that would be worth adding to your list:
It would be good to add "DISTINCT" to the group select statement... That
is, change (after my patch):
my $select= "SELECT $Attr->{grp_field}, $Attr->{uid_field}";
to:
my $select= "SELECT DISTINCT $Attr->{grp_
Thank you for your help, a workaround has been found.
The cause of failure seems to be using pnotes in a subroutine
called by the handler.
It is easy enough to work around the problem by avoiding the subroutine.
Forgive me if I've wasted the list's time with poor coding but perhaps
it is a bug. H
It should work fine. I wrote the same thing today (albeit without
method calls)...
# Trans handler
sub lookup_handler {
my $r=shift;
my $dbh=GTS::Util::connectdb(); # essentially a wrapper for DBI->connect
...
$r->pnotes(dbh=>$dbh);
return Apache2::Const::DECLINED;
}
# Response handler
-8<-- Start Bug Report 8<--
1. Problem Description:
I'm not sure if this is my interpretation of the documentation or a
real problem with the code. The documentation says:
> If the current request is a sub-request, this method returns a
>
Thanks Tom, good info I had not come across before! The page refers to
Apache 2, but they must have backported this to Apache 1.3 at some point
as well.
On Mon, 3 Apr 2006, Tom Schindl wrote:
Jonathan Field wrote:
Hi all,
I've noticed that under mod_perl any die() (or croak()) calls get
re
Geoff,
Sorry I'm not familiar with Apache2::TestModperl
search.cpan.org gives no hits for Apache2 TestModperl
How do I use this test to identify the problem with
MP2 pnotes?
I'm using
httpd-2.0.55
mod_perl-2.0.2
but can work with any trusted version.
If I comment out the line
# my $dbh = $r-
Thank you. Tried it, unfortunately no difference.
I'll try Geoffrey Young's test.
On Apr 3, 2006, at 12:08 PM, Torsten Foertsch wrote:
On Monday 03 April 2006 20:20, John Russell wrote:
sub handler($$) {
http://perl.apache.org/docs/2.0/user/porting/
porting.html#Method_Handlers
maybe that
John Russell wrote:
> In MP1 it was possible to do..
>
> Authorization Phase
>
> sub handler {
>
> my $dbh;
>
> $dbh = DBI->connect(..)
> || return "Error connecting.\n $DBI::errstr\n";
>
> $r->pnotes(DBH => $dbh);
>
>
> Response Phase
>
> sub handler($$) {
>
> my $dbh = $r
On Monday 03 April 2006 20:20, John Russell wrote:
> sub handler($$) {
http://perl.apache.org/docs/2.0/user/porting/porting.html#Method_Handlers
maybe that is your problem.
Torsten
pgpLWFPZiFN26.pgp
Description: PGP signature
Tom,
Thanks for your help.
I've only ever used pnotes to pass this DBI object so I've not had
experience
with other items. It worked fine in MP1.
I tried adding
use Apache2:RequestUtil();
to the existing entry in startup.pl
use Apache2::RequestRec ();
but it did not make any difference.
He
Is this only the case for DBI-Instances? I've been passing objects using
pnotes for quite a while in my apps and never had any problems.
What your the ... in your error message? Could you shown me the whole
code you are using maybe you forgot to load
Apache2::RequestUtil?
http://perl.apache.org/d
Tom,
under MP2
$r->pnotes(DBH => $dbh);
gives
failed for ..., reason: DBI::db=HASH(0x97c720c)
and also the same error with Apache2::SafePnotes.
Hence the reason to try de-referencing the hash.
Thank you, John.
On Apr 3, 2006, at 11:32 AM, Tom Schindl wrote:
John Russell wrote:
In MP1 it w
John Russell wrote:
> In MP1 it was possible to do..
>
> Authorization Phase
>
> sub handler {
>
> my $dbh;
>
> $dbh = DBI->connect(..)
> || return "Error connecting.\n $DBI::errstr\n";
>
> $r->pnotes(DBH => $dbh);
>
>
> Response Phase
>
> sub handler($$) {
>
> my $dbh = $r->
In MP1 it was possible to do..
Authorization Phase
sub handler {
my $dbh;
$dbh = DBI->connect(..)
|| return "Error connecting.\n $DBI::errstr\n";
$r->pnotes(DBH => $dbh);
Response Phase
sub handler($$) {
my $dbh = $r->pnotes('DBH');
However this doesn't work in MP2 because
Right - that was the line I was trying to find earlier. So much for my
theory about ignoring the LIMITs :-(
All I can think of to explain the speedup that people (including myself)
tend to see anyway is the indexes being cached in the key_buffer the
second+ times around.
Issac
Jeff wrote:
> -
If anything, it really doesn't make sense to cache something in the
query cache with limits intact; LIMIT is just a modifier which trims the
result set on the server side. Since LIMIT doesn't actually affect the
result set, per se, it doesn't make sense for the query cache to pay
attention to it a
23 matches
Mail list logo
