When a business puts up aÂ dedicated server, they want traffic. Traffic is the lifeblood of any internet business. But sometimes, too much traffic can seem like too much of a good thing. Too much traffic can cause a web server to crash and cause visitors to your server to be disappointed. Fortunately, there are ways to help websites that share content, such as Digg, to ”harden” your dedicated server so that it can handle those traffic spikes.
If you really need to handle a lot of hits in a short period of time, it is necessary to make sure that your web server is tweaked for maximum performance. It needs to be able to handle requests and deal with them as quickly as possible. There is always a tradeoff between memory used and resources available so this is why tweaking for your needs is so important.Â When you look at tweaking, you should start off by looking at what web server you are using.
There are several web servers out there to choose from. The Intel line includes a vast array of dedicated servers, usually offered with complete security. A professional web hosting company will ensure that the hardware used is up to date, efficient, managed and supported correctly. In addition, the web host should provide the solution that is best suited for your particular needs, with the best applications designed for your web site’s purpose.
That said, whatever solution you choose, most will require a level of management in order to use it well. Therefore, someone on your team needs to understand how to get the most out of it.
There is a lot of information out there about how to get the most out of a server in a situation where traffic is spiking. One of the easiest ways to tweak is to go into the configuration and set it for maximum number of clients. This will increase the memory load on it, but it will also make sure that people do not get an inaccessible site. This is not a foolproof way to handle a huge traffic spike, but in situations with huge spikes in traffic, it’s an integral part of a multifaceted approach to handle traffic spikes.
For those using MySQL on their servers, tweaking MySQL is another way to handle the Digg Effect. A slow MySQL server can be a real bottleneck on a server. Simple tweaking can be very beneficial, but it is best left to a professional who knows the finer points of your system and your web applications. Allowing more clients or caching more of the lookups can help improve response time, in general.
If your web application is using PHP then you should really be putting a PHP cache on your server. A PHP cache is a server application that keeps a copy of the PHP output of a script for future user. In practice, it might cache an output for anywhere from a few seconds to a few minutes. What this means is that if someone visits your page and then another person visits the page 2 seconds later, the first one will get the output produced by the PHP code, but the second visitor will get the output from the PHP cache. This means that code is executed once but used many times. For a dedicated server that gets low traffic, there isn’t much benefit, but for those servers that get huge traffic spikes, PHP caches can be very helpful.
There are several PHP caches available such as APC, eAccelerator, and mmcache to name just a few. APC is probably the best one to use for a dedicated server running PHP applications. It actually is slated to come out with PHP 6 and thousands of web servers are using it right now. Its installation is pretty straightforward and there are several great tutorials online for many different OS. Its configuration is easy and it doesn’t take much to get it going and begin to benefit from it.
While PHP caches are great for applications that run PHP, there are applications out there that do not run PHP so they can’t benefit from a PHP caches. There is a big gun out there that can give the benefits of a PHP cache for just about script or web app output. It’s called Squid. Squid is the granddaddy of caching proxy servers. Many people use it as a proxy server, but it does more than proxy, it can also cache the output of any page that it has visited. This means it can cache the output of a CGI script, it can cache the output of a Ruby script, it can even cache the output of a PHP script.
When utilized for server load management, a squid cache proxy server will stand between the end user and the web server. When a user wants to visit a page, they visit the cache proxy server and if the page is cached, it is sent to the end user. If it isn’t, then Squid will get the page from the web server and then send it to the end user. While it may seem a bit complicated, it’s completely transparent to the user. All they do is visit the web page and see the content.
Finally, there is a more radical approach to managing high traffic spikes on your dedicated server. This involves virtualization. Basically, you virtualize several installs. In one virtualization bucket you will have your web server. In another, you will have your mysql server, and in a still another will be the squid proxy cache server. This way, if one component is overwhelmed, the other components will be isolated. By breaking things up, you will have increased their overall efficiency.
Whichever methods you choose to ”Digg harden” your dedicated server, you can scale your server to handle your needs. You can rest assured that traffic will benefit your business and not hurt it productivity.
Superb Internet provides a complete range of managedÂ dedicated server web hosting solutions, from virtual shared web hosting through to high-volume, mission-critical dedicated server solutions. SuperbHosting.net’s core focus is customer service, and continuous service improvement, with offices and 24-hour/365-day operations in McLean, VA, Tukwila, WA, and Vancouver, BC, serving customers in over 150 countries worldwide.
View further information onÂ Dedicated Servers here.