Apache Mod-Python - mod_python trunk fails

Author

mod_python trunk fails

Jeff Hinrichs - DM&T

2006-05-21, 1:12 am

FreeBSD 6
Python 2.4.3

========================================
==============================
FAIL: test_phase_status (__main__.PerRequestTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
File "test.py", line 2127, in test_phase_status
self.fail(`rsp`)
AssertionError: '<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML
2.0//EN">\n<html><head>\n<title>500 Internal Server
Error</title>\n</head><body>\n<h1>Internal Server Error</h1>\n<p>The
server encountered an internal error or\nmisconfiguration and was
unable to complete\nyour request.</p>\n<p>Please contact the server
administrator,\n [no address given] and inform them of the time the
error occurred,\nand anything you might have done that may
have\ncaused the error.</p>\n<p>More information about this error may
be available\nin the server error log.</p>\n</body></html>\n'

----------------------------------------------------------------------
Ran 67 tests in 20.749s

FAILED (failures=1)
F Stopping Apache...
/usr/local/sbin/httpd -k stop -f
/home/jlh/mod_python/trunk/test/conf/test.conf

========================================
==============================
FAIL: testPerRequestTests (__main__.PerInstanceTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
File "test.py", line 2648, in testPerRequestTests
self.failUnless(result.wasSuccessful())
AssertionError

----------------------------------------------------------------------
Ran 6 tests in 62.204s

FAILED (failures=1)
[jlh@lakota:~/mod_python/trunk]$ svn info
Path: .
URL: http://svn.apache.org/repos/asf/httpd/mod_python/trunk
Repository Root: http://svn.apache.org/repos/asf
Repository UUID: 13f79535-47bb-0310-9956-ffa450edef68
Revision: 408123
Node Kind: directory
Schedule: normal
Last Changed Author: jgallacher
Last Changed Rev: 408048
Last Changed Date: 2006-05-20 12:44:51 -0500 (Sat, 20 May 2006)
Properties Last Updated: 2006-05-20 12:39:22 -0500 (Sat, 20 May 2006)

complete test run in attachment (mod_python.out)

This is my first failure report, so if I've included too much, too
little or incorrectly formatted the report please don't hesitate to
let me know.

-Jeff

--
Jeff Hinrichs
Dundee Media & Technology, Inc
jeffh@dundeemt.com
402.320.0821

Graham Dumpleton

2006-05-21, 1:12 am

For this particular failure, can you look in test/logs/error_log and
find section
which has:

[Mon May 08 21:58:55 2006] [error] phase_status_1
[Mon May 08 21:58:55 2006] [error] phase_status_2
[Mon May 08 21:58:55 2006] [error] phase_status_4
[Mon May 08 21:58:55 2006] [error] phase_status_5
[Mon May 08 21:58:56 2006] [error] phase_status_6
[Mon May 08 21:58:56 2006] [error] phase_status_7

You'll probably find not all lines are there due to error occurring.
Post to the
list the lines that were displayed and any error messages that follow
just
after that.

Nicolas did see this on Win32 amongst a lot of other failed tests. I
fixed the
other tests but didn't touch this one, yet the problem went away for
him when
he next tried the tests.

At the moment I suspect it may be something to do with Apache auth
config
setup for that test. Which version of Apache are you using??

Thanks.

Graham

On 21/05/2006, at 1:17 PM, Jeff Hinrichs - DM&T wrote:

> FreeBSD 6
> Python 2.4.3
>
> ========================================
==============================
> FAIL: test_phase_status (__main__.PerRequestTestCase)
> ----------------------------------------------------------------------
> Traceback (most recent call last):
> File "test.py", line 2127, in test_phase_status
> self.fail(`rsp`)
> AssertionError: '<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML
> 2.0//EN">\n<html><head>\n<title>500 Internal Server
> Error</title>\n</head><body>\n<h1>Internal Server Error</h1>\n<p>The
> server encountered an internal error or\nmisconfiguration and was
> unable to complete\nyour request.</p>\n<p>Please contact the server
> administrator,\n [no address given] and inform them of the time the
> error occurred,\nand anything you might have done that may
> have\ncaused the error.</p>\n<p>More information about this error may
> be available\nin the server error log.</p>\n</body></html>\n'
>
> ----------------------------------------------------------------------
> Ran 67 tests in 20.749s
>
> FAILED (failures=1)
> F Stopping Apache...
> /usr/local/sbin/httpd -k stop -f
> /home/jlh/mod_python/trunk/test/conf/test.conf
>
> ========================================
==============================
> FAIL: testPerRequestTests (__main__.PerInstanceTestCase)
> ----------------------------------------------------------------------
> Traceback (most recent call last):
> File "test.py", line 2648, in testPerRequestTests
> self.failUnless(result.wasSuccessful())
> AssertionError
>
> ----------------------------------------------------------------------
> Ran 6 tests in 62.204s
>
> FAILED (failures=1)
> [jlh@lakota:~/mod_python/trunk]$ svn info
> Path: .
> URL: http://svn.apache.org/repos/asf/httpd/mod_python/trunk
> Repository Root: http://svn.apache.org/repos/asf
> Repository UUID: 13f79535-47bb-0310-9956-ffa450edef68
> Revision: 408123
> Node Kind: directory
> Schedule: normal
> Last Changed Author: jgallacher
> Last Changed Rev: 408048
> Last Changed Date: 2006-05-20 12:44:51 -0500 (Sat, 20 May 2006)
> Properties Last Updated: 2006-05-20 12:39:22 -0500 (Sat, 20 May 2006)
>
> complete test run in attachment (mod_python.out)
>
>
> This is my first failure report, so if I've included too much, too
> little or incorrectly formatted the report please don't hesitate to
> let me know.
>
> -Jeff
>
> --
> Jeff Hinrichs
> Dundee Media & Technology, Inc
> jeffh@dundeemt.com
> 402.320.0821
> <mod_python.out>

Graham Dumpleton

2006-05-21, 7:11 am

On 21/05/2006, at 4:18 PM, Jeff Hinrichs - DM&T wrote:

> me culpa, forgot to CC the group
>
> ---------- Forwarded message ----------
> From: Jeff Hinrichs - DM&T <jeffh@dundeemt.com>
> Date: May 21, 2006 1:17 AM
> Subject: Re: mod_python trunk fails
> To: Graham Dumpleton <grahamd@dscpl.com.au>
>
>
> Graham,
> apache-2.2.2
>
> Here are the only lines that had the phrase "phase_status" I could
> find in the error_log:
> [Sun May 21 01:06:37 2006] [error] phase_status_1
> [Sun May 21 01:06:37 2006] [error] phase_status_2
> [Sun May 21 01:06:37 2006] [crit] [client 127.0.0.1] configuration
> error: couldn't check access. No groups file?: /tests.py

Hmmm, I thought I had worked this stuff out. :-(

Can you change test/test.py so the test configuration to that test
reads:

def test_phase_status_conf(self):
c = VirtualHost("*",
ServerName("test_phase_status"),
DocumentRoot(DOCUMENT_ROOT),
Directory(DOCUMENT_ROOT,
SetHandler("mod_python"),
AuthType("bogus"),
AuthName("bogus"),
Require("valid-user"),
PythonAuthenHandler
("tests::phase_status_1"),
PythonAuthenHandler
("tests::phase_status_2"),
PythonAuthenHandler
("tests::phase_status_3"),
PythonAuthzHandler
("tests::req_headers_out_access"),
PythonFixupHandler
("tests::phase_status_4"),
PythonFixupHandler
("tests::phase_status_5"),
PythonFixupHandler
("tests::phase_status_6"),
PythonHandler
("tests::phase_status_7"),
PythonDebug("On")))
return str(c)

Ie., add the PythonAuthzHandler entry.

Don't worry about what tests::req_headers_out_access is, I am just
using it
as a convenient function which simply returns apache.OK and does nothing
else.

Graham

Jeff Hinrichs - DM&T

2006-05-21, 1:11 pm

Graham,
I made the change,

Index: test.py
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
3D=3D=3D=3D
--- test.py (revision 408173)
+++ test.py (working copy)
@@ -2111,6 +2111,7 @@
PythonAuthenHandler("tests::phase_status=
_1"),
PythonAuthenHandler("tests::phase_status=
_2"),
PythonAuthenHandler("tests::phase_status=
_3"),
+
PythonAuthzHandler("tests::req_headers_out_access"),
PythonFixupHandler("tests::phase_status_=
4"),
PythonFixupHandler("tests::phase_status_=
5"),
PythonFixupHandler("tests::phase_status_=
6"),

The tests then ran successfully,
----------------------------------------------------------------------
Ran 67 tests in 21.310s

OK
Stopping Apache...
/usr/local/sbin/httpd -k stop -f
/home/jlh/mod_python/trunk/test/conf/test.conf
..
----------------------------------------------------------------------
Ran 6 tests in 62.728s

OK

-Jeff

On 5/21/06, Graham Dumpleton <grahamd@dscpl.com.au> wrote:
>
> On 21/05/2006, at 4:18 PM, Jeff Hinrichs - DM&T wrote:
>
>
> Hmmm, I thought I had worked this stuff out. :-(
>
> Can you change test/test.py so the test configuration to that test
> reads:
>
> def test_phase_status_conf(self):
> c =3D VirtualHost("*",
> ServerName("test_phase_status"),
> DocumentRoot(DOCUMENT_ROOT),
> Directory(DOCUMENT_ROOT,
> SetHandler("mod_python"),
> AuthType("bogus"),
> AuthName("bogus"),
> Require("valid-user"),
> PythonAuthenHandler
> ("tests::phase_status_1"),
> PythonAuthenHandler
> ("tests::phase_status_2"),
> PythonAuthenHandler
> ("tests::phase_status_3"),
> PythonAuthzHandler
> ("tests::req_headers_out_access"),
> PythonFixupHandler
> ("tests::phase_status_4"),
> PythonFixupHandler
> ("tests::phase_status_5"),
> PythonFixupHandler
> ("tests::phase_status_6"),
> PythonHandler
> ("tests::phase_status_7"),
> PythonDebug("On")))
> return str(c)
>
> Ie., add the PythonAuthzHandler entry.
>
> Don't worry about what tests::req_headers_out_access is, I am just
> using it
> as a convenient function which simply returns apache.OK and does nothing
> else.
>
> Graham
>

--=20
Jeff Hinrichs
Dundee Media & Technology, Inc
jeffh@dundeemt.com
402.320.0821

Graham Dumpleton

2006-05-22, 7:12 am

I have checked in a hopefully more permanent change to test harness to
cope with what you are seeing. I can only imagine it is something
strange
to do with what auth providers you have built into Apache.

On 22/05/2006, at 12:30 AM, Jeff Hinrichs - DM&T wrote:

> Graham,
> I made the change,
>
> Index: test.py
> ========================================
===========================
> --- test.py (revision 408173)
> +++ test.py (working copy)
> @@ -2111,6 +2111,7 @@
> PythonAuthenHandler
> ("tests::phase_status_1"),
> PythonAuthenHandler
> ("tests::phase_status_2"),
> PythonAuthenHandler
> ("tests::phase_status_3"),
> +
> PythonAuthzHandler("tests::req_headers_out_access"),
> PythonFixupHandler
> ("tests::phase_status_4"),
> PythonFixupHandler
> ("tests::phase_status_5"),
> PythonFixupHandler
> ("tests::phase_status_6"),
>
>
> The tests then ran successfully,
> ----------------------------------------------------------------------
> Ran 67 tests in 21.310s
>
> OK
> Stopping Apache...
> /usr/local/sbin/httpd -k stop -f
> /home/jlh/mod_python/trunk/test/conf/test.conf
> .
> ----------------------------------------------------------------------
> Ran 6 tests in 62.728s
>
> OK
>
> -Jeff
>
> On 5/21/06, Graham Dumpleton <grahamd@dscpl.com.au> wrote:
>
>
> --
> Jeff Hinrichs
> Dundee Media & Technology, Inc
> jeffh@dundeemt.com
> 402.320.0821

Graham Dumpleton

2006-05-23, 1:12 am

Jeff Hinrichs - DM&T wrote ..
> Graham,
> Would you like me to list the auth providers? It's built from the
> FreeBSD ports system. I've got PHP5 installed and vhosts.

Not at this time as am about to go on holidays for three weeks and have
got a major problem with some software at my work which is going to get
priority. In short my application didn't cope too well with connecting
to 150 different sites in one go. As far as I can tell it created a
message storm which the central application (Python wrappers around C++
messaging system), couldn't process before some connection liveness
timeouts fired which caused all the connections to be dropped and
started again. Thus it cause a never ending cycle. Lots of fun. :-)

Anyway, will probably look at it again after I get back and start looking
as implementing some auth provider hooks for mod_python.

I guess one thing you can do though is run the tests again and copy
out of the logs/error_log the sequence of phase_status_? messages
and post that back. On my box it still doesn't trigger the new one I
add corresponding to the authz phase, yet on your setup I assume it
is, I just don't actually know why it isn't triggering mine as Apache
code is such that it suggests it should. Thus, you box could be correct
and it is something about my build that is strange.

Graham

> On 5/22/06, Graham Dumpleton <grahamd@dscpl.com.au> wrote:
>
>
> --
> Jeff Hinrichs
> Dundee Media & Technology, Inc
> jeffh@dundeemt.com
> 402.320.0821

Jeff Hinrichs - DM&T

2006-05-23, 1:12 am

Graham,

Hope you have an enjoyable holiday. I am looking forward to holiday myself=
..

I'll get the info you requested off to you asap.

Will talk to you when you return.

-Jeff

On 5/22/06, Graham Dumpleton <grahamd@dscpl.com.au> wrote:
> Jeff Hinrichs - DM&T wrote ..
>
> Not at this time as am about to go on holidays for three weeks and have
> got a major problem with some software at my work which is going to get
> priority. In short my application didn't cope too well with connecting
> to 150 different sites in one go. As far as I can tell it created a
> message storm which the central application (Python wrappers around C++
> messaging system), couldn't process before some connection liveness
> timeouts fired which caused all the connections to be dropped and
> started again. Thus it cause a never ending cycle. Lots of fun. :-)
>
> Anyway, will probably look at it again after I get back and start looking
> as implementing some auth provider hooks for mod_python.
>
> I guess one thing you can do though is run the tests again and copy
> out of the logs/error_log the sequence of phase_status_? messages
> and post that back. On my box it still doesn't trigger the new one I
> add corresponding to the authz phase, yet on your setup I assume it
> is, I just don't actually know why it isn't triggering mine as Apache
> code is such that it suggests it should. Thus, you box could be correct
> and it is something about my build that is strange.
>
> Graham
>
o[vbcol=seagreen]
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
3D=3D=3D=3D=3D=3D=3D[vbcol=seagreen]
---[vbcol=seagreen]
---[vbcol=seagreen]
ld[vbcol=seagreen]
on[vbcol=seagreen]
>

--=20
Jeff Hinrichs
Dundee Media & Technology, Inc
jeffh@dundeemt.com
402.320.0821

Jeff Hinrichs - DM&T

2006-05-23, 1:13 pm

Graham,
I updated trunk (Revision: 408878).
Here are the desired lines from the error_log
------------------------------------------
[Tue May 23 07:18:28 2006] [notice] mod_python: (Re)importing module 'tests=
'
....skipping...
[Tue May 23 07:18:30 2006] [error] phase_status_1
[Tue May 23 07:18:30 2006] [error] phase_status_2
[Tue May 23 07:18:30 2006] [error] phase_status_4
[Tue May 23 07:18:30 2006] [error] phase_status_5
[Tue May 23 07:18:30 2006] [error] phase_status_6
[Tue May 23 07:18:30 2006] [error] phase_status_7
[Tue May 23 07:18:30 2006] [error] phase_status_8
[Tue May 23 07:18:30 2006] [error] phases =3D [1, 2, 5, 6, 7]
[Tue May 23 07:18:30 2006] [notice] mod_python: (Re)importing module
'mod_python.publisher'
[Tue May 23 07:18:30 2006] [notice] [client 127.0.0.1] Publisher
loading page /home/jlh/mod_python/trunk/test/htdocs/tests.py
---------------------------------------------------------------------------=
-----

-Jeff

On 5/22/06, Graham Dumpleton <grahamd@dscpl.com.au> wrote:
> Jeff Hinrichs - DM&T wrote ..
>
> Not at this time as am about to go on holidays for three weeks and have
> got a major problem with some software at my work which is going to get
> priority. In short my application didn't cope too well with connecting
> to 150 different sites in one go. As far as I can tell it created a
> message storm which the central application (Python wrappers around C++
> messaging system), couldn't process before some connection liveness
> timeouts fired which caused all the connections to be dropped and
> started again. Thus it cause a never ending cycle. Lots of fun. :-)
>
> Anyway, will probably look at it again after I get back and start looking
> as implementing some auth provider hooks for mod_python.
>
> I guess one thing you can do though is run the tests again and copy
> out of the logs/error_log the sequence of phase_status_? messages
> and post that back. On my box it still doesn't trigger the new one I
> add corresponding to the authz phase, yet on your setup I assume it
> is, I just don't actually know why it isn't triggering mine as Apache
> code is such that it suggests it should. Thus, you box could be correct
> and it is something about my build that is strange.
>
> Graham
>
o[vbcol=seagreen]
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
3D=3D=3D=3D=3D=3D=3D[vbcol=seagreen]
---[vbcol=seagreen]
---[vbcol=seagreen]
ld[vbcol=seagreen]
on[vbcol=seagreen]
>

--=20
Jeff Hinrichs
Dundee Media & Technology, Inc
jeffh@dundeemt.com
402.320.0821