The end goal of the Denial-of-service attack (DoS attack) or distributed denial-of-service attack (DDoS attack) is to make a resource unavailable to its users. The attack can be carried out on different levels, depending on the target and the people involved. In essence, the attack represents a highly intense attempt by one person or more to hinder websites or specialized internet services from running properly, causing financial losses to that resource's owner.

Depending on the motivation behind the attack, different targets might suffer from it. The attack can be implemented by a single person, carrying out a personal agenda. Such an attack is improbable to cause too much trouble and will only be temporary. A much bigger problem is organized attacks performed by a group of individuals in pursuit of financial gains. Such attacks can be sustained for longer periods of time and might as well ruin companies relying on online presence. Banks, online retailers, government sites, payment gateways, and even root nameservers can become easy prey for organized DDoS attacks.

"I think there is a world market for maybe five computers" is a famous misquote of Thomas J. Watson. The same Tom Watson, who took Computing Tabulating Recording Corporation and turned into what is today known as IBM. In essence, whoever said that the world will only need 5 computers wasn't totally wrong. Virtualization was first implemented in 1960 to aid better utilization of large mainframe hardware. 50 years later, it has grown to be part of all Fortune 100 success stories.

Lately, "fake servers" became a slightly satiric way to say "virtual servers". Why? Mainly because they are gaining a larger market share. Large enterprises were quick to adopt the virtual machines approach, while small business started late. However, by year-end 2010, enterprises with 100 to 999 employees will have a higher penetration of virtual machines deployed than the Global 500. For long years, small businesses could not afford even the entry level products, and this is what changed lately. Increased competition by server vendors has made server virtualization technology affordable to smaller companies.

You've seen 98.9%, 99.9%, and even 100% server uptime labels on hosting sites. The first thought which comes up to people is, "Well, 99.9% sounds pretty well; 0.01% is no big deal," and they are right. Many people are missing because even if the server is reachable, it doesn't necessarily mean it's fully functional. So many other things can go wrong with a server. A 99.9% uptime sticker is no guarantee that your business won't face different problems while the server is up and running. Hosting providers calculate uptime in ways that aren't intuitive from a user's perspective, not considering certain website downtime.

• First of all, not all downtime is counted. Scheduled maintenance is not part of the equation. Several hours of downtime is not what hosting companies would include. When monthly server uptime is calculated, such planned service interruptions are not included (deducted).
• Some, if not most, hosting providers tend to overlook shorter downtime periods. 3-5 minutes of unavailability are not considered significant and don't end up on the 99.9% sticker. You can see how multiple short downtime periods can make a difference at the end of the day.
• 0.01% means twice the trouble. 99.9% uptime means 8 hours and 45 minutes of downtime per year. So when you go down a notch to 98.9%, your calculated downtime goes up twice to 17 hours and 30 minutes.
• Resources are limited. "Unlimited" is just a nice way to sugarcoat "enough for the majority of common users." Your business might not be a common one, and demand might be high. There might be other significant sites sharing the same hosting hardware. At one point, it can all build up to an overload, resulting in poor loading times, broken transactions, and even a website. The server will remain online, but the performance won't meet your needs. Application monitoring is a good way to quickly and easily discover such problems.

100% uptime is not feasible. Eventually, something will go wrong. Network backbones will cause problems; power failures will happen, server software and hardware will fail now and then, and human error will always be important. So, my advice is: Don't try to find the ultimate provider, as it doesn't exist. Instead, find a decent provider with good server uptime and performance, 24-hour technical support, and frequent backups. Oh, why don't you go for a nice remote website monitoring service, just in case? It might be the best money you have ever spent.

Ping is one of the most well-known and easy-to-use computer network tools. Its basic application is to test the availability of hosts across an IP network. The ping tool was written by Mike Muuss way back in December 1983. Its practical use and reliability were undeniable until 2003, when a lot of internet service providers began filtering ICMP ping requests due to growing thread from internet worms and basic DDoS attacks. Nearly 3 decades later, ping stands tall, providing common computer users, webmasters, and network administrators with valuable information.

Check your network adapter

Common users can use ping from their local machine to check if their LAN, WLAN card is enabled and working by simply pinging itself. The command will look something very similar to this:

The response time of a webpage is the time elapsed from the moment the user requests a URL, until the request page is fully displayed. This process can be separated in to 3 logical units - transmission, processing and rendering.

• Transmission consists of the time to transmit the user's request and receive response from the server.
• Processing is the part of the operation where the server has to process the request and generate the response.
• Rendering is client side operation and consists of the time needed by the client machine to display the response to the user.