php.net |  support |  documentation |  report a bug |  advanced search |  search howto |  statistics |  random bug |  login
Bug #32897 session_start(): Failed to initialize storage module: user
Submitted: 2005-04-30 10:36 UTC Modified: 2005-05-06 03:32 UTC
From: dmih at in-solve dot ru Assigned:
Status: Not a bug Package: Session related
PHP Version: 4.3.11 OS: Win32/all
Private report: No CVE-ID: None
 [2005-04-30 10:36 UTC] dmih at in-solve dot ru
Description:
------------
This is repeat of lame "no-feedback" bug
http://bugs.php.net/bug.php?id=25876
with some new data. It seems that nobody will pay attention to the old bug with 'no feedback', so I have to make this new one with chance go give all needed feedback this time, because the bug is very annoying.

We are using PHP in virtual hosting environment, 1000+ hosts.
Apache 1.3 module.

We randomly see
session_start(): Failed to initialize storage module: user (path: d:\web\temp) on line 2.

where on line 2, 3, 4, etc there is only first any command that initialize session.

The problem is that the user NEVER set storage module to user! Just never.
Not in php.ini, not in scripts, nowhere.
We are hosting for 5 years now already and there is no errors in configuration.

No Zend, no accelerators/caches, just plain PHP with standard options.


I have feeling that this 'user' setting is coming from maybe some other execution task/thread. I also have feeling that one of apache working threads starts to always feel like 'user' session handler.

This seems to be related to this bug:
http://bugs.php.net/bug.php?id=32330
but situation may be more clear here because 'reproduction code' is surely shorter.

Reproduce code:
---------------
<?
 session_start();

Expected result:
----------------
Session startup

Actual result:
--------------
It works, but with some probability under heavy load:

session_start(): Failed to initialize storage module: user (path: d:\web\temp) on line 2.

Patches

Add a Patch

Pull Requests

Add a Pull Request

History

AllCommentsChangesGit/SVN commitsRelated reports
 [2005-04-30 10:46 UTC] dmih at in-solve dot ru
I said that bugs may be related to some apache working process thread.
I may add comment that will make clear why I feel so.
These bugs appearence pattern:

1. 10+ hours - apache working process is running, no bugs at all
2. something happens - I see these errors every 10 minutes or so.
3. apache working process restarts - still no bugs for 10+ hours.
 [2005-04-30 11:28 UTC] dmih at in-solve dot ru
It seems that this code fails to get proper handler setting.

	/* Open session handler first */
	if (PS(mod)->s_open(&PS(mod_data), PS(save_path), PS(session_name) TSRMLS_CC) == FAILURE) {
		php_error_docref(NULL TSRMLS_CC, E_ERROR, "Failed to initialize storage module: %s (path: %s)", PS(mod)->s_name, PS(save_path));
		return;
	}
 [2005-05-06 03:32 UTC] sniper@php.net
this is same as bug #32330, so don't reopen this anymore

 
PHP Copyright © 2001-2024 The PHP Group
All rights reserved.
Last updated: Sun May 12 10:01:31 2024 UTC