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.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
2 million visitors without any information on what you are hosting means nothing. Give more information so we can better suggest.
Teenager porn visitors from tumblr
I use 8 cpu 3.0ghz and 4gb ram ,about 20gb storage , it's wordpress blog
worldstream.nl - i do recommend you purchase the 1GBIT port upgrade, 100MBIT might be a bottleneck when theres high traffic.
Are you having issues with your current setup? If not, why change anything?
I have no issues but they don't have packages more that 8 cpu and when the website will grow i will not be enough
What is your CPU usage at currently? %
What is your RAM usage at out of 4GB?
It would be helpful if you answer @Awmusic question, it would make easier for people make sugestions.
If what you want is really stability than it woulb probably be better if you scale horizontally, and would be probably better to have DB on one server and your apache/nginx on another.
I don't understand on the server staff i need it all in one box
I use 6 cpu 100% and some time 8 cpu 100%
ram usage 4-5gb stable
That is roughly 1 visit per second (0.7 actually) - Install nginx somewhere on a box with a single CPU core and 1GB RAM, optimise MySQL a bit (max connections) and use PHP-FPM, then report back when you hit the 20million mark (which, with 7 visits per second, would be still cake for nginx, you need to worry about nginx at like 50000req/s).
Simply installing nginx won't help, usually the bottleneck is not the webserver, but the backend processing (like php).
I recommend HHVM or caching (varnish).
Nginx alone, with a default config, will easily speed it up while at the same time consume AT LEAST 50% less CPU/RAM than Apache. Switching to FPM (now likely modphp, maybe FCGI/FPM already though) will also reduce load.
I agree with @Aga If you are really using that much resources then you need to start scaling horizontally. A cluster approach would likely work much much better for your site. You need to separate your services into different servers and you DEFIANTLY need to use nginx+php-fpm if you are truly seeing that much traffic. Nginx will handle many more connections in the same resources than Apache will.
Cheers!
@William, yeah nginx seems like the next logical step from Apache. Read a lot about its scalability here http://highscalability.com/blog/2013/5/13/the-secret-to-10-million-concurrent-connections-the-kernel-i.html ... question, how does it not run out of ports?
HHVM isn't suitable for production use. He can simply install the apc pecl extension and that'll speed up PHP a lot. If he's using php 5.5 or newer then it has an opcode caching system built in and he can use that instead. He'll get 75% of the speed boost of HHVM but with php core level of stability.
If most of the pages on the site rarely change he can enable nginx fastcgi caching. This will make his site as fast as a static one and will reduce server resource usage as well. Varnish is not recommended because it is yet another thing that could break. There is no need for it because nginx can do caching too.
It does, but there is a fix for that in 1.9.1+ with linux kernel 3.9+ (so bleeding edge) as that integrated a kernel loadbalancing function for it which seems to perform pretty good.
https://www.nginx.com/blog/socket-sharding-nginx-release-1-9-1/
Hi, Looking for a sponsor?
What do you mean? You're asking the @OP if he wants a sponsorship or something?
Most likely he is asking for ad space or some kind of visibility in exchange for a server. Honestly, I wouldn't accept it if I were OP.
what about bandwidth? 10 TB? 20 TB? 40 TB?
I bid $50.00 for sponsorship.
@iContact get in touch with me on skype @readydedi.bharat and I'll suggest you a solution if you like we'll be delivering it to you.
Hello there,
Depending on your budget, we can offer you any of those packages with 24/7 Management included and up to 5 Gbps Anti DDoS Mitigation: https://www.hetnix.com/dedicated-servers.xml
Payable via PayPal | Card | Bitcoin | Bank.
If you need any additional info, feel free to contact us
Hi.
Shoot me a pm for review of your project. I can offer hosted solution with near 100% uptime (SLA is possible) with plus of performance and security improvements.
Hi,
What you need is more resources, so you can look for an enterprise VPS or upgrade to a Dedicated Server which has a lot more resources.
Since you're having a lot of traffic you need more processing and ram than a regular vps.
I got a high traffic WordPress blog which had a max of 3800 Online....And it handled it like a boss on a 4gb ram 4core Xeon E3 VPS. But the Admin Dashboard froze after crossing 2000 Online, but the frontend was blazing fast since all pages were cached in the RAM.
Answer to this is Use Nginx+Redis Cache - EasyEngine
With proper optimization, 2000k/day is quite easy. I'm running a forum on a Linode 4096 with 4 vCores, and it works like a charm under 2500k/day page requests including both read and write. For a wordpress site where most of the contents are static, 2000k/day shouldn't be that hard, not even talking about 2000k/month.
(If you are talking about 2000k/minute then that's a different story.)
woah... any tips? what stack do you use?
thanks.
How is reddis compared to memcached? My wordpress 1gb installs with apache+nginx from serverpilot with memcached is heavily swapping on digitalocean.
And it only get around 300 visitors a day.
Very interested in this discussion now.