8-bit characters and mod_perl/Apache-ASP

2001-05-10 Thread Carl Lipo 


Im having an odd (and annoying) problem with text data being returned via
Oracle::DBD (10.7) to an Apache::ASP (2.09) script. Essentially, perl that
returns perfectly valid characters in both cgi and as a .pl file fails to
return valid characters in the context of an Apache::ASP file. The issue,
I think has to do with the fact that the Oracle database (8.0.6) is
returning 8-bit characters. I can confirm this because I can repoint my
application to a version of the database set with 7bit characters and
everything works. While this shouldnt really be a problem (and isnt in the
normal 'perl' setting), Apache::ASP/mod_perl seems to be barfing on these
characters.

Here's some sample code:

use PNAP::Database::Query; ## a simplified database connection module
use CGI qw/:standard/;  ## for CGI output 
print header,
start_html('A Simple Example'),

## make a database connection
my $dbconn= new PNAP::Database::Query(Connection=NSPFinance, Debug=0);

## the sql statementA
my $orgSQL = select o.FULLNAME;
$orgSQL .=  from ORGANIZATION o, PARTY_ROLE pr, PARTY_ROLE_TYPE prt;
$orgSQL .=  where o.ORGANIZATIONOID = pr.SubjectPartyOID and;
$orgSQL .=  pr.PartyRoleTypeOID = prt.PartyRoleTypeOID and prt.ShortName = 
'NSP' and;
$orgSQL .=  pr.StartDate = sysdate and (pr.EndDate is null or pr.EndDate  
sysdate);
$orgSQL .=  order by FULLNAME;
my $rs = $dbconn-query(Sql=$orgSQL, UseCursor=0);
if ($rs)
{
 while ($rs-next())
 {
   my $fullname = $rs-FULLNAME; ## get the value from therecordset
   print Fullname: , $fullname,BR\n; ## output the value
  }
  $dbconn-finish($rs);
}


What is odd is that if I run this as a cgi (or from the command line), I
get this output:

Fullname: ATT
Fullname: Agis
Fullname: At Home
Fullname: CW
Fullname: CW/IDC
Fullname: Digex
Fullname: EBONE
...

However, the exact same code run as an Apache::ASP page results in this
output (on netscape for linux 4.76). 

Fullname: 
Fullname: 
Fullname: 
Fullname: 
Fullname: 
Fullname: 
...

On IE, however, I get this as output (viewed as source):

Fullname:  A T  T
Fullname:  A g i s
Fullname:  A t  H o m e
Fullname:  C  W
Fullname:  C W / I D C
Fullname:  D i g e x
Fullname:  E B O N E
...

This implies that the 8-bit characters are being rendered/dealt with
*differently* for the cgi version and the Apache::ASP version (and that
the Apache::ASP version honks up the 8-bit characters somehow). IE seems
to deal with the odd characters better than netscape. 

In order to address this problem, Ive tried all the standard approaches to
dealing with 8bit characters and Oracle databases (setting NLS_LANG, etc)
and have tried using Convert::Translit module to do the conversion..
However, none of those actions seem to make any difference. I still get
wonky output with the mod_perl/Apache::ASP version (but not the normal
Perl/cgi version). 

What I am wondering is:  

(1) is there a known problem with mod_perl and 8-bit character
sets? The version of apache I am running is 1.3.19 and mod perl 1.25, and
Apache 2.09.

(2) Is there any problem with this from the Apache::ASP side in either the
way the Response-Write is done or the way in which these character sets
are handled? Ive looked at the Response-Write and Response-WriteRef but
dont see anything obvious that could cause this.

(3) Has anyone seen this kind of behavior before?

Any help would be appreciated (it will beat having to export and re-import
the production database)...

Carl


Carl Lipo   EMAIL: [EMAIL PROTECTED]
IT Systems ArchitectPAGER: 206.570.2086
Internap Network Services   OFFICE: 206.504.5442
601 Union Street, Suite 1000FAX: 206.654.5684
Seattle, WA  98101  CELL: 206.579.0163

Unauthorized use, disclosure, or distribution of information within
this message is prohibited.

Re: Apache::ASP -- Corruption of statedir files/truncation of methodnames

2001-04-05 Thread Carl Lipo 


The statedir is local so it doesnt seem to be an NFS issue. There
definitely seems to be a corruption of the StateDir files: here is a error
that crops up in our error log.

[Thu Apr  5 08:09:07 2001] [notice] child pid 11555 exit signal
Segmentation fault (11)
Use of uninitialized value at /usr/lib/perl5/MLDBM.pm line 161 (#1)

(W) An undefined value was used as if it were already defined.  It was
interpreted as a "" or a 0, but maybe it was a mistake.  To suppress
this warning assign an initial value to your variables.



Unauthorized use, disclosure, or distribution of information within
this message is prohibited.

On Wed, 4 Apr 2001, Joshua Chamas wrote:

 Carl Lipo wrote:
  
  I've just started having problems with asp state files for Apache::ASP
  2.09 becoming corrupted during use. My application will work just fine but
  at somepoint the session asp state files get corrupted and only deleting
  them allows the application to continue working. Oddly, this occurs on two
  different servers so the problem does not appear to be machine specific.
  
  The symptom of the problem is that application errors are returned
  from the asp code execution that look something like this:
  
  Uncaught exception from user code:
  Can't locate object method 'Quer' via package
  quot;Apache::ASP::Request'; at (eval 263) a href=#167l
  ine 167/a. 
INAP::NEIS::_usr_local_applprod_infosource_apps_netcfg_viewNodes_htmxINL() called
  at /usr/local/lib/site_perl/Apache/
  ASP.pm line 1504
  
 
 Generally problems of this nature, where even perl data seems
 to be corrupted, seem to be from buffer overrun type issues
 that normally result in segfaults.  Looking at your config, 
 I would suspect the culprit to be use of DB_File on your
 StateDir, where /usr/local/etc/infosource/apps/netcfg/state-dir
 may be mounted over NFS.  NFS typically does not support the 
 flock() style locking that Apache::ASP uses to protect DB_File 
 from corruption.  
 
 If you are  must use NFS, don't use DB_File, use the default 
 SDBM_File, which does not corrupt as easily.  If you need to 
 get past the 1000 byte limit, then you might use 
 MLDBM::Sync::SDBM_File which I developed just for this purpose.
 I need to add config support for this still to Apache::ASP though.
 If you can switch to a CIFS/or samba mounted file system, do 
 this, as this supports flock() semantics.
 
 If StateDir is mounted locally, then possibly there is a bug in 
 Apache::ASP StateDB locking, but this is less likely the case
 and we might look at other types of buffer overrun issues that
 might corrupt perl data structures.  If your modperl httpd is compiled
 DSO, compile it static, as I have seen no end of odd errors with
 the DSO config from time to time.  
 
 Also, DBD::Oracle and long values with CLOBs/BLOBs can create 
 another type of buffer overrun, which is supposedly on OCI bug, 
 though you didn't mention use of Oracle this is an odd bug that 
 stung me recently.
 
 If you give up on trying to find this bug, which seems to be 
 related to StateDB corruption, you may create $Application  
 $Session objects with Apache::Session in the global.asa Script_OnStart, 
 which would at least eliminate the StateDB issue.
 
 -- Josh
 
 _
 Joshua Chamas Chamas Enterprises Inc.
 NodeWorks  free web link monitoring Huntington Beach, CA  USA 
 http://www.nodeworks.com1-714-625-4051

Apache::ASP -- Corruption of statedir files/truncation of methodnames

2001-04-04 Thread Carl Lipo 


I've just started having problems with asp state files for Apache::ASP
2.09 becoming corrupted during use. My application will work just fine but
at somepoint the session asp state files get corrupted and only deleting
them allows the application to continue working. Oddly, this occurs on two
different servers so the problem does not appear to be machine specific. 

The symptom of the problem is that application errors are returned
from the asp code execution that look something like this:

Uncaught exception from user code:
Can't locate object method 'Quer' via package
quot;Apache::ASP::Request'; at (eval 263) a href=#167l
ine 167/a. 
INAP::NEIS::_usr_local_applprod_infosource_apps_netcfg_viewNodes_htmxINL() called
at /usr/local/lib/site_perl/Apache/
ASP.pm line 1504

or

$Response::Appe not defined at /usr/local/lib/site_perl/Apache/ASP.pm a
href=#3132line 3132/a

Where the object method that cant be located is truncated from the actual
ASP method (i.e., Query will be truncated to be Quer or $Response::Append
to log will be tructed at Response::Appe). What appears to be happnening
is that the namespace is corrupted somehow and the Apache::ASP/mod_perl is
not mapping the names of methods correctly. 

When we delete the state files, the problem goes away (and then returns at
some point in the future). Has anyone encountered this problem? Is there a
bug in the 2.09 code?

This is our environment:
Linux,  Apache 1.3.19, mod_perl 1.25
Apache::ASP 2.09


Here is the actual error output:

table
trtd valign=topfont size=-1b Global: /btd valign=topfont size=-1 
/usr/local/etc/infosource/apps/netcfg/
trtd valign=topfont size=-1b   File: /btd valign=topfont size=-1 
/usr/local/applprod/infosource/apps/netcfg/viewNodes.htm
trtd valign=topfont size=-1b IP: /btd valign=topfont size=-1 
64.94.117.16
trtd valign=topfont size=-1b HTTP_*: /btd valign=topfont size=-1 
HTTP_ACCEPT= image/gif, image/x-xbitmap, image/jpeg, image/pjpeg, image/png, */*br
  HTTP_ACCEPT_CHARSET= iso-8859-1,*,utf-8br
  HTTP_ACCEPT_ENCODING= gzipbr
  HTTP_ACCEPT_LANGUAGE= enbr
  HTTP_CONNECTION= Keep-Alivebr
  HTTP_COOKIE= session-id=0eef76d5377c6e62def6333423014869br
  HTTP_HOST= applprod4.sea.internap.combr
  HTTP_REFERER= https://applprod4.sea.internap.com/apps/netcfg/searchNode.htmbr
  HTTP_USER_AGENT= Mozilla/4.75 [en] (X11; U; Linux 2.2.16 i686)
trtd valign=topfont size=-1b  Query: /btd valign=topfont size=-1 
trtd valign=topfont size=-1b   Form: /btd valign=topfont size=-1 search= 
  Search  br
  searchLocation= br
  searchNodeName= br
  searchNodeStatus= br
  searchNodeType= br
  searchRack= br
  searchSiteCode= seabr
  sessionActive= Active
/table

tt
buErrors Output/u/b
ol

li Uncaught exception from user code:
Can't locate object method quot;Querquot; via package 
quot;Apache::ASP::Requestquot; at (eval 263) a href=#167line 167/a.
INAP::NEIS::_usr_local_applprod_infosource_apps_netcfg_viewNodes_htmxINL() 
called at /usr/local/lib/site_perl/Apache/ASP.pm line 1504
eval {...} called at /usr/local/lib/site_perl/Apache/ASP.pm line 1504
Apache::ASP::Execute('Apache::ASP=HASH(0x8d7d930)') called at 
/usr/local/lib/site_perl/Apache/ASP.pm line 174
Apache::ASP::handler('Apache=SCALAR(0x8d2b978)') called at /dev/null line 0
eval {...} called at /dev/null line 0
, /usr/local/lib/site_perl/Apache/ASP.pm line 1506
/ol

buDebug Output/u/b
ol

li Uncaught exception from user code:
Can't locate object method quot;Querquot; via package 
quot;Apache::ASP::Requestquot; at (eval 263) line 167.
INAP::NEIS::_usr_local_applprod_infosource_apps_netcfg_viewNodes_htmxINL() 
called at /usr/local/lib/site_perl/Apache/ASP.pm line 1504
eval {...} called at /usr/local/lib/site_perl/Apache/ASP.pm line 1504
Apache::ASP::Execute('Apache::ASP=HASH(0x8d7d930)') called at 
/usr/local/lib/site_perl/Apache/ASP.pm line 174
Apache::ASP::handler('Apache=SCALAR(0x8d2b978)') called at /dev/null line 0
eval {...} called at /dev/null line 0
, /usr/local/lib/site_perl/Apache/ASP.pm line 1506
/ol
/tt
pre

buASP to Perl Script/u/ba name=1nbsp;/a

a name=2  1/a: package INAP::NEIS; ;; sub 
INAP::NEIS::_usr_local_applprod_infosource_apps_netcfg_viewNodes_htmxINL {  ;;  @_ = 
(); ;; no strict;;use vars qw($Application $Session $Response $Server $Request);;use 
lib qw(/usr/local/etc/infosource/apps/netcfg/);;;
a name=3  2/a: 
a name=4  3/a: $Response-gt;AppendToLog(quot;[NEIS] - init.htm:quot;);
a name=5  4/a: 
a name=6  5/a: 
a name=7  6/a: ###
a name=8  7/a: #
a name=9  8/a: #init.htm
a name=10  9/a: #
a name=11 10/a: #   Dexter and Wayland
a name=12 11/a: #
a name=13 12/a: #   October 2000
a name=14 13/a: #
a name=15 14/a: #   10/12/00  Wayland Wasserman
a name=16 15/a: #   Common header file for entire NCM app
a name=17 16/a: #   Contains database

Re: Apache::ASP -- Corruption of statedir files/truncation of methodnames

2001-04-04 Thread Carl Lipo 


In addition:  here are the httpd.conf configurations for this application:

FilesMatch "\.(htm|asp)$"
PerlSetVar Global /usr/local/etc/infosource/apps/netcfg/
# PerlSetVar Global .
PerlSetVar GlobalPackage INAP::NEIS
PerlSetVar StateDir /usr/local/etc/infosource/apps/netcfg/state-dir
PerlSetVar CookiePath  /apps/netcfg
SetHandler perl-script
PerlHandler Apache::ASP Apache::SSI
PerlSetVar SessionTimeout 90
#0 production, 2 or 1 for dev
PerlSetVar Debug 0
PerlSetVar StatINCMatch PNAP|NCM
# PerlSetVar StatINC 1
#0 production, text compression
# PerlSetVar DynamicIncludes 0
PerlSetVar Clean 0
PerlSetVar StateDB DB_File
PerlSetVar AllowSessionState 1
PerlSetVar DebugBufferLength 500
PerlSetVar StateCache 0
# PerlSetVar UseStrict 1
PerlSetVar TimeHiRes 1
/FilesMatch


On Wed, 4 Apr 2001, Carl Lipo wrote:

 
 I've just started having problems with asp state files for Apache::ASP
 2.09 becoming corrupted during use. My application will work just fine but
 at somepoint the session asp state files get corrupted and only deleting
 them allows the application to continue working. Oddly, this occurs on two
 different servers so the problem does not appear to be machine specific. 
 
 The symptom of the problem is that application errors are returned
 from the asp code execution that look something like this:
 
 Uncaught exception from user code:
 Can't locate object method 'Quer' via package
 quot;Apache::ASP::Request'; at (eval 263) a href=#167l
 ine 167/a. 
INAP::NEIS::_usr_local_applprod_infosource_apps_netcfg_viewNodes_htmxINL() called
 at /usr/local/lib/site_perl/Apache/
 ASP.pm line 1504
 
 or
 
 $Response::Appe not defined at /usr/local/lib/site_perl/Apache/ASP.pm a
 href=#3132line 3132/a
 
 Where the object method that cant be located is truncated from the actual
 ASP method (i.e., Query will be truncated to be Quer or $Response::Append
 to log will be tructed at Response::Appe). What appears to be happnening
 is that the namespace is corrupted somehow and the Apache::ASP/mod_perl is
 not mapping the names of methods correctly. 
 
 When we delete the state files, the problem goes away (and then returns at
 some point in the future). Has anyone encountered this problem? Is there a
 bug in the 2.09 code?
 
 This is our environment:
 Linux,  Apache 1.3.19, mod_perl 1.25
 Apache::ASP 2.09
 
 
 Here is the actual error output:
 
 table
 trtd valign=topfont size=-1b Global: /btd valign=topfont size=-1 
/usr/local/etc/infosource/apps/netcfg/
 trtd valign=topfont size=-1b   File: /btd valign=topfont size=-1 
/usr/local/applprod/infosource/apps/netcfg/viewNodes.htm
 trtd valign=topfont size=-1b IP: /btd valign=topfont size=-1 
64.94.117.16
 trtd valign=topfont size=-1b HTTP_*: /btd valign=topfont size=-1 
HTTP_ACCEPT= image/gif, image/x-xbitmap, image/jpeg, image/pjpeg, image/png, */*br
   HTTP_ACCEPT_CHARSET= iso-8859-1,*,utf-8br
   HTTP_ACCEPT_ENCODING= gzipbr
   HTTP_ACCEPT_LANGUAGE= enbr
   HTTP_CONNECTION= Keep-Alivebr
   HTTP_COOKIE= session-id=0eef76d5377c6e62def6333423014869br
   HTTP_HOST= applprod4.sea.internap.combr
   HTTP_REFERER= https://applprod4.sea.internap.com/apps/netcfg/searchNode.htmbr
   HTTP_USER_AGENT= Mozilla/4.75 [en] (X11; U; Linux 2.2.16 i686)
 trtd valign=topfont size=-1b  Query: /btd valign=topfont size=-1 
 trtd valign=topfont size=-1b   Form: /btd valign=topfont size=-1 
search=   Search  br
   searchLocation= br
   searchNodeName= br
   searchNodeStatus= br
   searchNodeType= br
   searchRack= br
   searchSiteCode= seabr
   sessionActive= Active
 /table
 
 tt
 buErrors Output/u/b
 ol
 
 li Uncaught exception from user code:
   Can't locate object method quot;Querquot; via package 
quot;Apache::ASP::Requestquot; at (eval 263) a href=#167line 167/a.
   INAP::NEIS::_usr_local_applprod_infosource_apps_netcfg_viewNodes_htmxINL() 
called at /usr/local/lib/site_perl/Apache/ASP.pm line 1504
   eval {...} called at /usr/local/lib/site_perl/Apache/ASP.pm line 1504
   Apache::ASP::Execute('Apache::ASP=HASH(0x8d7d930)') called at 
/usr/local/lib/site_perl/Apache/ASP.pm line 174
   Apache::ASP::handler('Apache=SCALAR(0x8d2b978)') called at /dev/null line 0
   eval {...} called at /dev/null line 0
 , /usr/local/lib/site_perl/Apache/ASP.pm line 1506
 /ol
 
 buDebug Output/u/b
 ol
 
 li Uncaught exception from user code:
   Can't locate object method quot;Querquot; via package 
quot;Apache::ASP::Requestquot; at (eval 263) line 167.
   INAP::NEIS::_usr_local_applprod_infosource_apps_netcfg_viewNodes_htmxINL() 
called at /usr/local/lib/site_perl/Apache/ASP.pm line 1504
   eval {...} called at /usr/local/lib

[ANNOUNCE] HTTP::WebTest 0.01 released to CPAN

2001-01-13 Thread Carl Lipo 


HTTP::WebTest (by Richard Anderson) was released to CPAN today and is
available for download. It is a module for creating automated unit tests
for Apache::ASP pages (and other). Here is the announcement posted to
comp.lang.perl.announce and comp.lang.perl.modules:

NAME
HTTP::WebTest - Test remote URLs or local web files

DESCRIPTION
This module runs tests on remote URLs or local web files containing
Perl/HTML/JavaScript/etc. and generates a detailed test report. The
test specifications can be read from a parameter file or input as
method arguments. If you are testing a local file, Apache is
started on a private/dynamic port with a configuration file in a
temporary directory.  The module displays the test results on the
terminal by default or directs them to a file. The module will also
optionally e-mails the test results. When the calling program
exits, the module stops the local instance of Apache and deletes
the temporary directory.

Each test consists of literal strings or regular expressions that
are either required to exist or forbidden to exist in the fetched
~page. You can also specify tests for the minimum and maximum number
of bytes in the returned page. If you are testing a local file, the
module checks the error log in the temporary directory before and
after the file is fetched from Apache. If messages are written to
the error log during the fetch, the module flags this as an error
and writes the messages to the output test report.

SYNOPSIS
 This module can accept input parameters from a parameter file or
 subroutine arguments.

 TO RUN WEB TESTS DEFINED BY SUBROUTINE ARGUMENTS:

 use HTTP::WebTest; run_web_test(\@web_tests, \$num_fail,
\$num_succeed, \%test_options)

 or

 use HTTP::WebTest; run_web_test(\@web_tests, \$num_fail,
\$num_succeed)

 TO RUN WEB TESTS DEFINED BY A PARAMETER FILE:

 use sigtrap qw(die normal-signals); # Recommended, not necessary
 use HTTP::WebTest; $webtest = HTTP::WebTest-new();
 $webtest-web_test('my_web_tests.wt', \$num_fail, \$num_succeed);

 The web_test() method has an option to test a local file by
 starting Apache on a private port, copying the file to a temporary
 htdocs directory and fetching the page from Apache.  If you are
 testing with multiple parameter files, you can avoid restarting
 Apache each time by calling new() only once and recycling the
 object:

 use sigtrap qw(die normal-signals); # Recommended, not necessary
 use HTTP::WebTest;
 $webtest = HTTP::WebTest-new();
 foreach $file (@ARGV) {
$webtest-web_test($file, \$num_fail, \$num_succeed);
 }

 TO ENABLE DEBUGGING MESSAGES (OUTPUT TO STDOUT):

 If you are calling the web_test method, use the debug parameter.
 If you are calling the run_web_test method, do this:

 use HTTP::WebTest;
 $HTTP::WebTest::Debug = 1; # Diagnostic messages
 $HTTP::WebTest::Debug = 2; # Messages and preserve temp Apache dir
 run_web_test(\@web_tests, \$num_fail, \$num_succeed)

RESTRICTIONS / BUGS
This module only works on Unix (e.g., Solaris, Linux, AIX, etc.).
The module's HTTP requests time out after 3 minutes (the default
value for LWP::UserAgent). If the file_path parameter is specified,
Apache must be installed. If the file_path parameter is specified,
the directory /tmp cannot be NFS-mounted, since Apache's lockfile
and the SSL mutex file must be stored on a local disk.

VERSION
This document describes version 0.01, release date 13 January 2001.

TODO
Add option to validate HTML syntax using HTML::Validator. Add
option to check links (see
http://world.std.com/~swmcd/steven/perl/pm/lc/linkcheck.html).

AUTHOR
 Richard Anderson [EMAIL PROTECTED]

COPYRIGHT
Copyright (c) 2000 Richard Anderson. All rights reserved. This
module is free software. It may be used, redistributed and/or
modified under the terms of the Perl Artistic License.

[EMAIL PROTECTED]  RayCosoft, LLC
Perl/Java/Oracle/Unix software engineeringwww.unixscripts.com
www.zipcon.net/~starfire/home Seattle, WA, USA

Apache::ASP question: undefined subroutines

2000-10-30 Thread Carl Lipo 


Does anyone know what causes these kinds of errors? The 'subroutine' that
the output is referring to is the name of the page (in this case
search.htm) that is being loaded and that sits in the content directory --
there are no real missing subroutines...Maddeningly, if I hit reload
enough times the page will finally load at some point and then error a few
reloads later.

The page (search.htm) sits on an NFS mount (could that be a problem?). 

Errors Output 

1.Undefined subroutine
InterNAP::NFT::_home_staff_carl_htdocs_apps_nft_search_htmxINL called at
/usr/local/lib/site_perl/Apache/ASP.pm line 1395. ,
   /usr/local/lib/site_perl/Apache/ASP.pm line 1397 

Debug Output 

1.Undefined subroutine
InterNAP::NFT::_home_staff_carl_htdocs_apps_nft_search_htmxINL called at
/usr/local/lib/site_perl/Apache/ASP.pm line 1395. ,
   /usr/local/lib/site_perl/Apache/ASP.pm line 1397 

Here is the ASP Cconfiguration information for the app:

   Directory /office/common/infosource/apps/nft
FilesMatch "\.(htm|asp)$"
SetHandler perl-script
PerlHandler Apache::ASP Apache::SSI
PerlSetVar Filter On
PerlSetVar IncludesDir .
PerlSetVar StateDB DB_File
PerlSetVar CookiePath /apps/nft
PerlSetVar Global .
PerlSetVar StateDir /tmp/.state-asp/nft
PerlSetVar GlobalPackage InterNAP::NFT
PerlSetVar AllowSessionState 1
PerlSetVar SessionTimeout 60
PerlSetVar Debug 2
PerlSetVar MailHost mailhost-stu.sea.internap.com
PerlSetVar MailErrorsTo [EMAIL PROTECTED]
PerlSetVar CONFIG_FILE /usr/local/etc/nft/nft.xml
/FilesMatch
/Directory

Any ideas would be greatly appreciated -- its so intermittent, its
driving us crazy!

----
Carl Lipo   EMAIL: [EMAIL PROTECTED]
IT Systems ArchitectPAGER: 888.393.5620
InterNAP Network Services   OFFICE: 206.504.5442
601 Union Street, Suite 1000FAX: 206.264.1833
Seattle, WA  98101  CELL: 206.579.0163

Unauthorized use, disclosure or distribution of information within
this message is prohibited.

Re: Apache::ASP and clock times

2000-07-07 Thread Carl Lipo 


We are not sure why but it appears that the time problem may end up being
specific to a DSO enabled, debian distribution of apache. When we
recompiled a static version, the time that the ASP processes reported
remained the same as the system clock (no changes from PDT to GMT). It
looks like this is a bad-compilation problem or an effect of DSO. 

On Fri, 7 Jul 2000, Vivek Khera wrote:

  "JC" == Joshua Chamas [EMAIL PROTECTED] writes:
 
 JC It may be possible that the time calculation is based
 JC on some TZ* environment variable that is not consistent,
 
 It has been my experience that the first time any of the system
 date/time routines are called and TZ is set, that becomes the
 permanent time zone until the process goes away.  So it could be
 something totally other than ASP setting the time zone for the
 process.

Apache::ASP and clock times

2000-07-05 Thread Carl Lipo 



Carl Lipo   EMAIL: [EMAIL PROTECTED]
IT Systems ArchitectPAGER: 888.393.5620
InterNAP Network Services   OFFICE: 206.504.5442
601 Union Street, Suite 1000FAX: 206.264.1833
Seattle, WA  98101  CELL: 206.579.0163

Unauthorized use, disclosure or distribution of information within
this message is prohibited.

Re: Apache::ASP and clock times

2000-07-05 Thread Carl Lipo 



 That the error_log date itself is changing seems
 to me that Apache also thinks its the wrong time.
 If this were consistent, I would think that your box
 just has the wrong timezone set, but that this is
 inconsistent is bizarre.

Yes, this is the problem. If it was just a matter of being one time or the
other, we could deal. But since its inconsistent -- its a serious
problem.
  
 It may be possible that the time calculation is based
 on some TZ* environment variable that is not consistent,
 perhaps being set by some scripts  something?  Still
 bizarre but seems plausible to me.  I would log the 
 data from %ENV in Script_OnStart , and see how it differs 
 between these time changes.

Thats what we thought. However, if we print out the TZ variable in the
Script_OnStart it consistently comes out PDT regardless of the time that
the server reports. Argh.

 
 Anyone else have any ideas here ?
 
 Looking at your Apache::ASP settings I would recommend
 that you set StateDir to some /tmp or /var area, instead
 of the default in Global/.state for better housekeeping.
 Global is really good for includes, global.asa, and 
 perl modules, and I would keep the state files out of 
 there which are dynamic bits.
 

Good point. We need to clean that up.

 --Joshua
 
 Carl Lipo wrote:
  
  We are having a problem with our Apache::ASP scripts and the clock time
  they are reporting. It seems that when an ASP script is run, the internal
  clock is sometimes set to GMT. An example from our error log is shown
  below. The problem is that our scripts expect localtime (PDT).  The other
  odd (and particularly frustrating) thing is that this problem is *not*
  consistent. Sometimes we get GMT while other times we get localtime (PDT).
  Its very odd. We have narrowed the problem down to mod_perl and/or
  Apache::ASP. Has anyone seen this behavior before? Is there a
  configuration problem?
  
  The problem seems to start with the ScriptOnStart (not before that...) --
  and lasts until the end of the ASP processing.
  
   This is PDT, the correct local time 
  [Wed Jul  5 16:10:56 2000] [error] [asp] [29715] [debug] ASP object
  created -
  Application: Apache::ASP::Application=HASH(0x83532a8); GlobalASA: 
Apache::ASP::GlobalASA=HASH(0x81df93c); Internal: Apache::ASP::State=HASH(0x81eab5c); 
Request: Apache::ASP::Request=HASH(0x87ed180); Response: 
Apache::ASP::Response=HASH(0x87ed198); Server: Apache::ASP::Server=HASH(0x848891c); 
Session: Apache::ASP::Session=HASH(0x81e97a8); app_state: 1; basename: index.htm; 
buffering_on: 1; cgi_headers: 0; clean: 0; compile_includes: 0; cookie_path: /; dbg: 
2; debugs_output: ARRAY(0x81a6990); errs: 0; filehandle: GLOB(0x8353284); filename: 
/office/common/infosource-wally/index.htm; filter: 1; global: /etc/apache/state-asp; 
global_package: ; group_refresh: 120; id: 
_office_common_infosource_wally_index_htmxINL; includes_dir: .; init_packages: 
ARRAY(0x8488964); mail_alert_to: ; mail_errors_to: [EMAIL PROTECTED]; mtime: 
962475992; no_cache: ; no_session: ; no_state: ; package: 
Apache::ASP::Compiles::_etc_apache_state_asp_global_asa; paranoid_session: 0; 
pod_comments: 1; r!
:
  Apache=SCALAR(0x81e9610); remote_ip: 216.223.5.149; secure_session: ; 
session_serialize: ; session_timeout: 1200; stat_inc: ; stat_inc_match: ; 
state_cache: ; state_db: DB_File; state_dir: /etc/apache/state-asp/.state; 
state_manager: 10; ua: Mozilla/4.72
  [en] (X11; U; Linux 2.2.13 i686); unique_packages: 0;
  [Wed Jul  5 16:10:56 2000] [error] [asp] [29715] [debug] inlining include 
home-go.include
  [Wed Jul  5 16:10:56 2000] [error] [asp] [29715] [debug] inlining include 
links/internap.links
  [Wed Jul  5 16:10:56 2000] [error] [asp] [29715] [debug] active undefing
  sub 
Apache::ASP::Compiles::_etc_apache_state_asp_global_asa::_office_common_infosource_wally_index_htmxINL
  code CODE(0x87ef78c) before compiling
  [Wed Jul  5 16:10:56 2000] [error] [asp] [29715] [debug] compiling into
  package Apache::ASP::Compiles::_etc_apache_state_asp_global_asa subid
  
Apache::ASP::Compiles::_etc_apache_state_asp_global_asa::_office_common_infosource_wally_index_htmxINL
  
  
  -- note here is the Script_OnStart. Notice how the time changes!
  
  [Wed Jul  5 23:10:56 2000] [error] [asp] [29715] [debug] Script_OnStart
  [Wed Jul  5 23:10:56 2000] [error] [asp] [29715] [debug] executing
  Script_OnStart
  [Wed Jul  5 23:10:56 2000] [error] [asp] [29715] Time is now (localtime(time()):
  [Wed Jul  5 23:10:56 2000] [error] [asp] [29715] 56 10 23 5 6 100 3 186
   printing localtime(time()) [above] shows that perl really thinks it
   is this later [GMT time] 
  
  [Wed Jul  5 23:10:56 2000] [error] [asp] [29715] [debug] executing
  _office_common_infosource_wally_index_htmxINL
  [Wed Jul  5 23:10:58 2000] [error] [asp] [29715]  
INDEX.HTM*
  [Wed Jul  5 23:10:58 2000] [error] [asp] [29715] * INDEX USER: carl*
  [Wed Jul  5 23:10:59 2000] [error] [asp] [29715] [debug] active undefing

ASP.pm and Date::Manip/DBD::Sybase

1999-12-20 Thread Carl Lipo 


I am having recurring troubles using the Date::Manip and DBD::Sybase
modules in the contest of mod_perl/ASP pages. 75%-90% of the time my ASP
pages will work just fine.  The reminder 25-10% I get these kinds of
errors:

[Fri Dec 17 12:27:51 1999] [error] [asp] [7922] [error] ERROR: Date::Manip
unable to determine TimeZone. --   Date::Manip::Date_TimeZone called at
/usr/lib/perl5/site_perl/Date/Manip.pm line 608 -- Date::Manip::Date_Init()
called at /usr/lib/perl5/site_perl/Date/Manip.pm line 1295 --
Date::Manip::ParseDate('today') called at (eval 49) line 103 --

or

The context allocation routine failed when it tried to load localization
files!! One or more following problems may caused the failure
Cannot access the sybase home directory, please check environment variable
SYBASE or ~sybase [Fri Dec 17 15:51:57 1999] [error] [asp] [16617] [error] 
install_driver(Sybase) failed: DBD::Sybase initialize: cs_ctx_alloc()
failed at/usr/lib/perl5/i586-linux/5.00404/DynaLoader.pm line 185. --
at /usr/lib/perl5/site_perl/PNAP/Database/Query.pm line 276 --
(in cleanup) Can't call method "disconnect" without a package or
object reference at /usr/lib/perl5/site_perl/PNAP/Database/Query.pm line
294. -- , /usr/lib/perl5/site_perl/Apache/ASP.pm line 1102

The odd thing is that at some times the problems are worse than others. I
havent been able to correlate this behavior with memory or cpu usage. I
tried everything I can think of to stop this problem: I have tried using a
PerlRequire entry in the httpd.conf to include the modules when the server
starts up. I have added the two ENV variables to the httpd.con as
PerlSetVar and PerlPassEnv entries. I explicitely use BEGIN statements in
the ASP code and set the ENV variables. However, nothing so far has worked
-- at some frequency at least *some* of the children cannot get the TZ or
the SYBASE environmental variable. Its driving me nuts!

In the archives, I saw that at least one or two people in the past noted
this problem but I never saw how it was resolved. Is there something I'm
missing in the mod_perl configuration? Something I need to do? Any advice?

The server is Intel Linux 2.0.35 
Server version: Apache/1.3.9 (Unix)
Perl 5.00503
mod_perl Version 1.20
ASP version 0.16 (but also has same behavior with 0.17)

----
Carl Lipo   EMAIL: [EMAIL PROTECTED]
IT Systems ArchitectPAGER: 888.393.5620
InterNAP Network Services   OFFICE: 206.504.5442
601 Union Street, Suite 1000FAX: 206.264.1833
Seattle, WA  98101  CELL: 206.579.0163