SugarCRM causes apache2 to crash on "Load Module" page

[wdroush]

Sugar Version: 5.2, 5.5RC (tried both, same issue)
Sugar Edition Sugar Community
Category: Modules
Operating System: Ubuntu 9.04
PHP Version: 5.2.6
Database: MySQL 5.0.75
Web Server: Apache 2.2.11

Steps to reproduce
1. Click "Load Modules" from admin menu
2. Everything will load fine.
3. Click anywhere else, Apache process is aready dead from the end of the last page call.

Only the load modules page does this, only this application, we host many applications and haven't had this issue before.


Error message or issue presented
Apache Log:
[Fri Oct 16 08:09:22 2009] [error] [client 172.16.100.184] File does not exist: /var/www/internal/crm/include/javascript/yui/ext/resources, referer: http://172.16.100.20/crm/index.php?m...rd&view=module
[Fri Oct 16 08:09:22 2009] [error] [client 172.16.100.184] File does not exist: /var/www/internal/crm/include/javascript/yui/ext/resources, referer: http://172.16.100.20/crm/index.php?m...rd&view=module
[Fri Oct 16 08:09:22 2009] [error] [client 172.16.100.184] File does not exist: /var/www/internal/crm/include/javascript/yui/ext/resources, referer: http://172.16.100.20/crm/index.php?m...rd&view=module
[Fri Oct 16 08:09:22 2009] [error] [client 172.16.100.184] File does not exist: /var/www/internal/crm/include/javascript/yui/assets, referer: http://172.16.100.20/crm/index.php?m...rd&view=module
[Fri Oct 16 08:09:22 2009] [error] [client 172.16.100.184] File does not exist: /var/www/internal/crm/include/javascript/yui/assets, referer: http://172.16.100.20/crm/index.php?m...rd&view=module
[Fri Oct 16 08:09:36 2009] [notice] caught SIGTERM, shutting down

^Me restarting Apache, it isn't responding.

SugarCRM Log:
Reports nothing.

I know it looks like an Apache issue, but why only this application, this page, and errors are reported with these scripts?

I've tried fresh installs of SugarCRM, both the stable and RC, no difference.

Thanks.


Edit:
3-0 19129 34/103/103 W 0.82 396 0 57.0 0.14 0.14 172.16.100.184 Server.local GET /crm/index.php?module=Administration&action=index HTTP/1.1

That is the data of the process in apache that is frozen. Execution time is way over 300 seconds, weird eh?

[wdroush]

Current "workaround" is to open the load module page, restart apache2, upload file, restart apache2, install file, and go back out to the main menu. Modules install fine this way.

It's only this page, and I shouldn't be doing too much with it, HOWEVER we host production websites, and while a second of being down isn't a huge deal, it does bother me, and especially if another administrator doesn't remember to stay out of the Load Module section and causes the CRM to freeze until I go in and restart the process.

[wdroush]

Any ideas?

[eggsurplus]

I'd recommend finding those missing files from an install zip and move those files over to production so that they don't come up missing anymore.

[wdroush]

I'd recommend finding those missing files from an install zip and move those files over to production so that they don't come up missing anymore.

They aren't there, I've tried the two different versions too.

[eggsurplus]

Those directories exist in my version (5.2.0j) and have existed in past versions. I wonder if it creates those directories on the fly? If so, could it be a file permissions issue?

[wdroush]

Fixed the missing files, still crashing apache, now there are no logs on either files.

[wdroush]

PHP.info
http://www.certifiedfinancialpg.com/phpinfo.htm



Apache site:

<VirtualHost 172.16.100.20:80>
ServerAdmin webmaster@localhost

DocumentRoot /var/www/internal
<Directory "/">
		Options FollowSymLinks
		AllowOverride AuthConfig

</Directory>
<Directory /var/www/internal>
	Options Indexes FollowSymLinks MultiViews
	AllowOverride all
	Order allow,deny
        Allow from 127.0.0.0/255.0.0.0 ::1/128
        Allow from 172.16.100.0/255.255.255.0
</Directory>
ScriptAlias /cgi-bin/ /usr/lib/cgi-bin/
<Directory "/usr/lib/cgi-bin">
	AllowOverride None
	Options +ExecCGI -MultiViews +SymLinksIfOwnerMatch
	Order allow,deny
	Allow from all
</Directory>
ErrorLog /var/log/apache2/error.log

# Possible values include: debug, info, notice, warn, error, crit,
# alert, emerg.
LogLevel warn

CustomLog /var/log/apache2/access.log combined

   Alias /doc/ "/usr/share/doc/"
   <Directory "/usr/share/doc/">
       Options Indexes MultiViews FollowSymLinks
        AllowOverride None
        Order deny,allow
        Deny from all
        Allow from 127.0.0.0/255.0.0.0 ::1/128
    </Directory>
</VirtualHost>

Any ideas? I think I can bump the number of processes per page (err site, whatever), but that will only work until all the processes are locked. I could write a script to trash bad threads, but I shouldn't have to, etc. etc.

[wdroush]

It is killing the virtual host, and only the virtual host... all other sites work, but no other scripts on the same virtual host... any ideas what would cause that?

[wdroush]

Bingo

Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Including generic hook file
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Retrieve UpgradeHistory : SELECT upgrade_history.* FROM upgrade_history  WHERE upgrade_history.id = '78a9b5a6-6958-e51a-b26a-4adcf008da45'
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Limit Query:SELECT upgrade_history.* FROM upgrade_history  WHERE upgrade_history.id = '78a9b5a6-6958-e51a-b26a-4adcf008da45'  Start: 0 count: 1
Mon Nov  2 13:13:12 2009 [26428][1][INFO] Query:SELECT upgrade_history.* FROM upgrade_history  WHERE upgrade_history.id = '78a9b5a6-6958-e51a-b26a-4adcf008da45'  LIMIT 0,1
Mon Nov  2 13:13:12 2009 [26428][1][INFO] Query Execution Time:0.000111103057861
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Including generic hook file
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] rebuilding cache for Administration
Mon Nov  2 13:13:12 2009 [26428][1][INFO] Query:SELECT * FROM fields_meta_data WHERE  custom_module='Administration' AND  deleted = 0
Mon Nov  2 13:13:12 2009 [26428][1][INFO] Query Execution Time:0.000227212905884
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Including generic hook file
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Retrieve UpgradeHistory : SELECT upgrade_history.* FROM upgrade_history  WHERE upgrade_history.id = '433fac49-ad01-d718-c7b0-4ad8afca3954'
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Limit Query:SELECT upgrade_history.* FROM upgrade_history  WHERE upgrade_history.id = '433fac49-ad01-d718-c7b0-4ad8afca3954'  Start: 0 count: 1
Mon Nov  2 13:13:12 2009 [26428][1][INFO] Query:SELECT upgrade_history.* FROM upgrade_history  WHERE upgrade_history.id = '433fac49-ad01-d718-c7b0-4ad8afca3954'  LIMIT 0,1
Mon Nov  2 13:13:12 2009 [26428][1][INFO] Query Execution Time:0.000112056732178
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Including generic hook file
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] rebuilding cache for Administration
Mon Nov  2 13:13:12 2009 [26428][1][INFO] Query:SELECT * FROM fields_meta_data WHERE  custom_module='Administration' AND  deleted = 0
Mon Nov  2 13:13:12 2009 [26428][1][INFO] Query Execution Time:0.000231027603149
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Including generic hook file
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Retrieve UpgradeHistory : SELECT upgrade_history.* FROM upgrade_history  WHERE upgrade_history.id = 'da329ee0-b199-f46f-e997-4ad8c86588ee'
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Limit Query:SELECT upgrade_history.* FROM upgrade_history  WHERE upgrade_history.id = 'da329ee0-b199-f46f-e997-4ad8c86588ee'  Start: 0 count: 1
Mon Nov  2 13:13:12 2009 [26428][1][INFO] Query:SELECT upgrade_history.* FROM upgrade_history  WHERE upgrade_history.id = 'da329ee0-b199-f46f-e997-4ad8c86588ee'  LIMIT 0,1
Mon Nov  2 13:13:12 2009 [26428][1][INFO] Query Execution Time:0.000291109085083
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Including generic hook file
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] rebuilding cache for Administration
Mon Nov  2 13:13:12 2009 [26428][1][INFO] Query:SELECT * FROM fields_meta_data WHERE  custom_module='Administration' AND  deleted = 0
Mon Nov  2 13:13:12 2009 [26428][1][INFO] Query Execution Time:0.000109910964966
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Including generic hook file
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Retrieve UpgradeHistory : SELECT upgrade_history.* FROM upgrade_history  WHERE upgrade_history.id = '165159e2-7e6e-d742-aff6-4adcc74ad9b1'
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Limit Query:SELECT upgrade_history.* FROM upgrade_history  WHERE upgrade_history.id = '165159e2-7e6e-d742-aff6-4adcc74ad9b1'  Start: 0 count: 1
Mon Nov  2 13:13:12 2009 [26428][1][INFO] Query:SELECT upgrade_history.* FROM upgrade_history  WHERE upgrade_history.id = '165159e2-7e6e-d742-aff6-4adcc74ad9b1'  LIMIT 0,1
Mon Nov  2 13:13:12 2009 [26428][1][INFO] Query Execution Time:0.000113964080811
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] Including generic hook file
Mon Nov  2 13:13:12 2009 [26428][1][DEBUG] USING HTTPS TO CONNECT TO HEARTBEAT

Freezes after that.

Accepting all traffic on the firewall gets rid of the issue.

Weird, why is SugarCRM doing HTTPS requests on this page?