Regular Server crash

[AdelM]

Hi I'm encountering big peroblems with a fresh installed plesk server, updated from plesk 9.3.0 to 9.5.1 then 9.5.2 running debian Lenny 5 and PHP Version 5.2.6-1
Version Parallels Plesk Panel v9.5.2_build95100504.17 os_Debian 5.0
OS Linux 2.6.18-028stab068.9


Looking into /var/log/sw-cp-server/error_log :
I have the same error : [ERROR]: spawning fcgi failed

Any help would be much appreciated.

Best Regards

----------------------------------------------------
2010-08-23 23:10:02: (log.c.75) server started
2010-08-23 23:10:17: (mod_fastcgi.c.1748) connect failed: Cannot allocate memory on unix:/opt/psa/tmp/sw-engine.sock-0
2010-08-23 23:10:17: (mod_fastcgi.c.2897) backend died, we disable it for a 5 seconds and send the request to another backend instead: reconnects: 0 load: 1
2010-08-23 23:10:19: (mod_fastcgi.c.2834) socket failed: Cannot allocate memory 10 1024
2010-08-23 23:12:03: (log.c.135) server stopped
2010-08-23 23:12:04: (log.c.75) server started
2010-08-23 23:12:57: (mod_fastcgi.c.2834) socket failed: Cannot allocate memory 10 1024
2010-08-23 23:12:59: (mod_fastcgi.c.865) failed: Cannot allocate memory
2010-08-23 23:12:59: (mod_fastcgi.c.1105) [ERROR]: spawning fcgi failed.
2010-08-23 23:12:59: (mod_fastcgi.c.3505) all handlers for /index.php on .php are down.
2010-08-23 23:13:13: (mod_fastcgi.c.2834) socket failed: Cannot allocate memory 10 1024
2010-08-23 23:13:16: (mod_fastcgi.c.865) failed: Cannot allocate memory
2010-08-23 23:13:16: (mod_fastcgi.c.1105) [ERROR]: spawning fcgi failed.
2010-08-23 23:13:16: (mod_fastcgi.c.3505) all handlers for /index.php on .php are down.
2010-08-23 23:13:29: (mod_fastcgi.c.865) failed: Cannot allocate memory
2010-08-23 23:13:29: (mod_fastcgi.c.1105) [ERROR]: spawning fcgi failed.
2010-08-23 23:13:34: (log.c.135) server stopped
2010-08-23 23:13:35: (log.c.75) server started
2010-08-23 23:17:07: (mod_fastcgi.c.1748) connect failed: Cannot allocate memory on unix:/opt/psa/tmp/sw-engine.sock-0
2010-08-23 23:17:07: (mod_fastcgi.c.2897) backend died, we disable it for a 5 seconds and send the request to another backend instead: reconnects: 0 load: 1
2010-08-23 23:17:07: (log.c.135) server stopped
2010-08-23 23:17:08: (log.c.75) server started
2010-08-23 23:17:09: (mod_fastcgi.c.2834) socket failed: Cannot allocate memory 10 1024
2010-08-23 23:17:16: (mod_fastcgi.c.865) failed: Cannot allocate memory
2010-08-23 23:17:16: (mod_fastcgi.c.1105) [ERROR]: spawning fcgi failed.
2010-08-23 23:17:16: (mod_fastcgi.c.3505) all handlers for /index.php on .php are down.
2010-08-23 23:17:22: (mod_fastcgi.c.865) failed: Cannot allocate memory
2010-08-23 23:17:22: (mod_fastcgi.c.1105) [ERROR]: spawning fcgi failed.
2010-08-23 23:27:11: (log.c.135) server stopped
2010-08-23 23:27:12: (log.c.75) server started

 

[ideasmultiples]

Probably need more memory on your server: Cannot allocate memory on unix:/opt/psa/tmp/sw-engine.sock-0

How much memory you have now?

 

[AdelM]

Hi

Thanks for your reply

It's a virtual server with 2 Gb of memory (burstable up to ?); a previous server with less memory was running normally and never crashed, can't understand.

 

[IgorG]

You should check resources from Virtuozzo side if it is Virtuozzo VPS.

 

[AdelM]

Hi

I do not have acces to virtuozzo (hosting company restriction), is there another way to check this ?

 

[IgorG]

You can check output of free command at least. If it is UBC (not SLM) configured VPS, you can check failcounters with

# cat /proc/user_beancounters

Anyway only provider can increase resources for your VPS.

 

[AdelM]

I asked my provider and he increased my ressources. He answred that I've exceeded the number of allowed TCP connections, I have 20 websites on this server most of them have max 50 visits/day two websites generates up to 300 visits/day. Can it be something maliciuous !!??