6

configuring APACHE for high traffic

view full story
linux-howto

http://serverfault.com – I've started getting high volume traffic to me server (unmanaged server, contos 64bit 32G RAM with 2G throughput). It seems that the server status is 99% idle, the throughput is way less then 2G but the apache keeps chocking. I tried changing the following in the WHM - Apache configuration: **Minimum Spare Servers [?] 5 default** **Maximum Spare Servers [?] 20** **Server Limit (maximum: 20,000) [?] 2000** **Max Clients [?] 2000** but nothing helps...it freezes in: ps ax|grep httpd | wc -l : 806 and my websites are unreachable or timeout, or very very very slow... PLEASE - HELP!! t (HowTos)

taskli's picture
Submitted by taskli on 06/26/2012 – Made popular on 06/26/2012

Stories similar to configuring APACHE for high traffic

Hello my server goes down when apache connections rise about 250. I set Server Limit and Max Clients in whm (Global Configuration) Server Limit = 2000 Max Clients = 2000

also I set this in pre_virtualhost_global.conf (Include Editor) ServerLimit 2000 MaxClients 2000

I think it should be ok now because I see about 2000 DOTs id apache status page: Server uptime: 5 hours 55 minutes 38 se

We have a very high configuration server which is holding 2 websites with considerably good traffic. In peak hours we have 3000 to 4000 simultaneous users. But the current configuration of our apache seems to be wrong, as it hangs or makes the site very slow in the peak hours.

I would like to create a high-available setup (e.g. a small cluster) for a webserver, i.e. it will run Apache, PHP and MySQL.

There will be between 2-8 small websites running with only very little traffic and workload.

We have WebLogic installed on two separate RHEL 5.6 servers. An Apache web server instance will be installed on each of those servers which will be configured to listen on 80 and 443 and redirect any requests from 80 to 443. SSL will terminate at the Apache web server and traffic between the Apache web server and the WebLogic servers will not need to be encrypted.

I've scoured the web but I can't find out what MaxSpareServers are in Apache MPM prefork..

The MaxSpareServers directive sets the desired maximum number of idle child server processes. An idle process is one which is not handling a request.

I have a memory problem with my servers (Debian 6.0) where is installed Apache httpd 2.2, mod_php and APC.

Actually it seems that httpd allocate all available memory (4GB) when is waiting too much for an answer from Mongodb (30 seconds timeout).

Apache Traffic Server is one of the most admired services to have come under the Apache Server Foundation project. Traffic Server is essentially a caching and load balancing server that manages your traffic in an effective manner.

Maybe I'm going about this the wrong way, but it seems like it would be a good idea to limit maximum requests per second on a per IP address basis for an Apache server.

I've seen the results before (over-limit warnings), especially for things like APIs which can easily get flooded with requests.

We recently moved our Wordpress blog to a dedicated server (Apache) that already served ~4,000 pages daily. Yesterday the server crashed. We had a high volume of visitors to the blog, ~4,000 with a peak of 500 around the crash.

The weird thing is, the server does not time out, it just crashes and we have to manually restart it.