httpd crashes = website is inacessible

[Junta Momonari]

Hi !

I have a problem with my website on a virtual server with Plesk (9.0.1, but the problem was also present before the update [8.2 to 9.0.1]).

The website is sometime temporarily inaccesible (the pages are not in 404, they are loading indefinitly and so they are not displayed).
I found that a restart of the httpd service solved this bug and after the website is again accessible.
When I do not restart the service, the website becomes accessible alone after approximately 1 or 2 hours.

I dont know very well the administration of this type of server...

Do you have an idea and/or a solution to resolve my bug ?

Thanks.

 

[atomicturtle]

what distro are you on, and how much ram do you have? That sounds to me like you're running out of resources

 

[onsightdata]

What do your logs show?

Also do you have watchdog running? and if so does it alert you that the web server crashed? You will need to provide a few more details on what you checked so far. If you are unsure how to check logs and stuff it might be best to open a support ticket with your datacenter or parallels and a tech could look into the issue for you .

 

[Junta Momonari]

what distro are you on, and how much ram do you have? That sounds to me like you're running out of resources

Version de Parallels Plesk Panel 9.0.1

It's a virtual server on this machine :
Système d'exploitation Linux 2.6.9-023stab048.6-smp
CPU AuthenticAMD, Dual-Core AMD Opteron(tm) Processor 2216 HE

For my server :
Memory guaranteed : 256 MB
Dynamic memory : 1 GB

Two screns that I made a few days ago when I spoke about this bug with my colleagues.
http://img10.imageshack.us/img10/4384/chgsrvx.jpg
http://img17.imageshack.us/img17/9897/chgsrvv.jpg

The traffic of the website did not increase this last year, so I have elimitate the possibility of saturation of my server (but the real machine...).

A precision : the bugs are juste present on the morning. The afternoom we never had this problem...

What do your logs show?

Excuse me, but wich log ? Error log ?
Do I must search a specific message ?

Also do you have watchdog running? and if so does it alert you that the web server crashed? You will need to provide a few more details on what you checked so far.

No. I don't know "wathdog".

If you are unsure how to check logs and stuff it might be best to open a support ticket with your datacenter or parallels and a tech could look into the issue for you .

I try, but my hoster is stupid (to say in a polite).
He replies to me that it's a software problem and so it's not his problem (to him, just hardware problem are supported)...
I just obtained some information : to him, no problem on the lan flow an no problem on the real server.

 

[Junta Momonari]

Hello!

Since my previous post I found and installed watchdog.
This highlights regular bugs of the Apache server.
At the same time I observed in the services manager, some alerts in the black zone of "numproc" in parallel with these bugs of Apache.

An idea to address the parameters of the "numproc"?

Thank you.

 

[Faris Raouf]

numproc is the number of processes.

Basically you need a more powerful VPS. Memory and disk space are not the only things that differentiate a VPS.

Hopefully your provider will be able to tell you if the next most powerful option they provide also has a higher numproc.