Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


Which CPU model good for handle very large traffic amount website
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

Which CPU model good for handle very large traffic amount website

ziditzidit Member
edited October 2015 in General

Hello,

Currently, I have dedicated server with E3-1281v3 cpu to serve 1000k+ a day (Pageview) for 4 websites (wordpress). Server is overload for some period when concurrent visitor very high (2000 - 4000 concurrent) I can see number of processes that being processed 1000+ processes which result in server hang until restart.

When it hang

So, I am looking for more powerful CPU model which can handle this type of traffic very well. (I do not need to monitor them 24/7 for restarting thing)

Now, I have some cpu model in my mind. Please suggest me if they can handle this amount of traffic or I need to find more powerful model.

http://www.cpubenchmark.net/compare.php?cmp[]=2051&cmp[]=2516&cmp[]=2389

Between E5-2620v2 and E5-1650v3. From my point, e5-1650v3 win with CPU mark score But E5-2620v2 have more core? Which likely to handle more concurrency process?

Please suggest :)

«1

Comments

  • Have you done any optimizations to the configuration or do you just want a bigger box?

    Are you using caching on the wordpress sites?

  • @wych said:
    Have you done any optimizations to the configuration or do you just want a bigger box?

    Are you using caching on the wordpress sites?

    I have done many of optimization to reduce load, caching, cloudflare, etc.

  • you should be looking at geekbench not cpubenchmark

  • @zidit said:
    I have done many of optimization to reduce load, caching, cloudflare, etc.

    Have you tried HHVM ?

    http://nomadhacker.com/2014/02/10/whip-snapping-fast-wordpress-nginx-hhvm-part-2/

  • To many processes.
    And concurrent users is way to low to be bogged down by that CPU. Just to low.

    Something isn't right. I'd look at what's using the most CPU and troubleshoot from there.

    WordPress has plugins that are horribley inefficient. I'd look at what's being used.

    Not uncommon to host 100's of WordPress on a single CPU server. If CO figured correctly.

  • sinsin Member
    edited October 2015

    @zidit said:
    I have done many of optimization to reduce load, caching, cloudflare, etc.

    Install W3TC (or Supercache) and set it to generate/serve static pages to your visitors (or use fastcgi cache with a Wordpress purge plugin) so php shouldn't be called at all unless you're in wp-admin or preloading cache...you honestly shouldn't have any problems with that amount of traffic on a decent VPS.

    Thanked by 1zidit
  • @TarZZ92 said:
    you should be looking at geekbench not cpubenchmark

    Sorry, but why? :) I never check geekbench before. Please advise

  • zidit said: Sorry, but why? :) I never check geekbench before. Please advise

    better info, various results, and realistic

    just type the CPU in the box http://browser.primatelabs.com/

    ideally for single core look for 3000+ anything lower will be quite bad for what you require. multicore 13000+

    Thanked by 1zidit
  • Try moving one of your sites to an easyengine install using nginx cache & hhvm, as long as your site isn't too dynamic, you should be able to handle boat loads of traffic. https://rtcamp.com/easyengine/

    I don't think there's any reason your current server couldn't handle that load on a well cached site.

    Thanked by 1zidit
  • I have upload image of "top" when it hang for more information

  • Never seen a load of 1000+ and didn't though it was possible also...

    Thanked by 1zidit
  • MunMun Member

    Yeah that is configuration issues. What Web server??

  • @FredQc said:
    Never seen a load of 1000+ and didn't though it was possible also...

    Yes, it has happened but it drastically load reduced immediately after httpd restart

  • @TarZZ92 said:

    Many thanks :) I am checking

  • Not yet, I am not geek guy for these things but I will try. My server using cPanel. Is there any problem for pluging this with cpanel?

  • 2000 visitors a second? 1000000 visitors a day gives me 11.57 hits a second, which should be no problem for any well optimized site.

    Could this be a HTTP flood ddos attack? check your access log or analytics and see how much is real traffic

  • @linuxthefish said:
    2000 visitors a second? 1000000 visitors a day gives me 11.57 hits a second, which should be no problem for any well optimized site.

    Could this be a HTTP flood ddos attack? check your access log or analytics and see how much is real traffic

    It's real traffic. These websites just have a lot of visitor

  • @Mun said:
    Yeah that is configuration issues. What Web server??

    I use cPanel / apache

  • smansman Member
    edited October 2015

    Is that all Apache using 0.5%-0.6% each adding up to all that CPU usage?

    This may be a dumb question but do you have KeepAlive enabled?

    In httpd.conf

    KeepAlive On

    Have you tried other obvious optimizations like reducing MaxClients and increasing MaxRequestsPerChild?

  • cfgguycfgguy Member, Host Rep

    You need load balancer with 2 web server and 1 database server.

  • cfgguycfgguy Member, Host Rep

    And what is the network port like ??

  • ziditzidit Member
    edited October 2015

    @cfgguy said:
    And what is the network port like ??

    100mbps Now, I utilize 60-80mbps. Based on graph from provider.

  • @sman said:
    Your screen capture doesn't show us what is using all your CPU resources. The web processes (apache?) are hardly using any and MySQL is only using 29%. Your are using 86.9% of 400% = 346%. So what is using the other 317%? Is that all Apache using 0.5%-0.6% each adding up to all that?

    This may be a dumb question but do you have Keepalive enabled?

    In httpd.conf

    KeepAlive On

    Keepalive = On

    Most of process is php and httpd

  • Time to load balance horizontally (i.e. multiple machines). Setup a load balancer, a set of web servers and a mysql cluster.

  • Hi.

    I'd recommend to hire a sysadmin (can suggest myself). It's rather easy to handle the load with WordPress.

  • MunMun Member

    @Profforg said:
    Hi.

    I'd recommend to hire a sysadmin (can suggest myself). It's rather easy to handle the load with WordPress.

    Agreed especially with that hardware.

  • If you are on cpanel try lite speed free trial and enable LS cache it is 10 minute 3 line install drop in replacement... You will not recognize your server

    https://www.litespeedtech.com/support/wiki/doku.php?id=litespeed_wiki:cpanel:auto-installer

    and enable cache

    https://www.litespeedtech.com/support/wiki/doku.php/litespeed_wiki:cache:lscache:setup-guidline

    then hire a sysadmin

  • @Mun said:
    Agreed especially with that hardware.

    Again I agree it needs optimization. What he is doing is nothing in that hardware. A atom should be able to run that.....

  • 8 way e7 8890v3 with 12 TB of ram.

    Thanked by 1yomero
  • risharderisharde Patron Provider, Veteran

    geezus! 1000+ load! I've had a website waaaaay back years ago that took a million hits per month and my shared provider never complained. I'm guessing as the rest of the guys are saying, this has to be some sort of configuration problem and looks like apache from the screenshot.

    Fix configs on apache
    Cache optimizations will help if you don't have any cache enabled
    If these are logged in users, you might want to also look at optimizing mysql as well since mysql performance affects your page load times

    Really interested in hearing the outcome of if you were able to optimize this since I'm going to be doing something big with a server soon as well

    Thanked by 1zidit
Sign In or Register to comment.