Understanding the Method of Making Money Through Web-hosting Service

Web hosting can be considered as a subscription oriented business. Once a customer subscribes to your hosting service, you will then earn constant money, on a monthly basis and that too with no additional investment from your side. If you don’t have the right quantity of money towards a web hosting business, then in that case, you can start a reseller web hosting business for a far lesser price. You will then comfortably be able to offer the reseller hosting package, along with the whole seller free and domain name registration facility. On the other hand, investing in a web hosting business through the traditional way will involve huge investment whereby you would be investing big time money towards purchasing the equipment and software. Hence, starting a reseller web hosting business is the right choice.

Starting a reseller web hosting business is a better option to earn money. A reseller hosting account involves a particular web hosting plan, by which the reseller can buy the hosting service at a discount or at a wholesale rate and then offer the same hosting service to a third party. Resellers of web hosting plans are nothing but entrepreneurs who act as middleman. Resellers of web hosting plans basically lease their hosting service from another web host’s server rather than taking the pains to purchase and maintain their own equipment. By this way, a seller saves on basic investment over heads. This is because a reseller who deals with web hosting service, don’t have to worry about maintaining equipment or software. In case there arise any problem with the server, it is the whole seller who will resolve the issue, not the re-seller. The customer will not be aware as to who actually restored the server issue.

A reseller web hosting business can be conducted from anywhere around. The basic criteria required for a web hosting business is the internet access. If you want to be successful in your web hosting business, you need to think local. Check for new customers who are around you. In other words, look out for customers at your workplace, your neighborhood, friends and relatives too. You could also find customers from your cell phone numbers and email contact list. This is a simple idea of fishing for customers who are within your source of contact. You can promote your web hosting business by posting on relevant blogs, forums and using text link ads. You can send your company related information through news letters to targeted customers. You can also have a website, containing various sections, in the form of advertisements. Your visitor will then select on the appropriate advertising section of interest and thereby get access to more information about the concerned web hosting package and thus be tempted to buy what they need. By this method, you are actually helping your visitors to narrow down their search and be able to select a package that is suitable for their business.

A reseller web hosting business can be started by a person who knows basic windows experience. You don’t have to be an expert in java developer or any kind of software programming expert. When you enter into the reseller web hosting business, it is important to know the reliability and the credibility of the company. Otherwise, you can be in losses and ultimately end up in subsequent closure of your reseller hosting account. I would like to conclude this article by encouraging you to take up a re-seller web hosting business which involves a very small investment plan. The investment is in the form of a monthly or yearly fee for availing this opportunity of conducting the concerned reseller web hosting business.

The Casablanca Method of Server Monitoring

Server Monitoring tools can capture such a wide variety of server metrics that many find it difficult to determine what metrics they should monitor.

Actually, actor Claude Rains gave us the answer back in 1942. In the movie “Casablanca”, Rains’s character “Captain Renault” announces “Round up the usual suspects” — and that’s exactly where your server monitoring should begin!

Who are the usual suspects? Simple enough — the main components of our server are the CPU, Memory, and the Disk. Applications can become CPU bound, Memory bound, or Disk bound — so we should collect metrics that reflect this.

Our CPU metric usual suspects should be:

Processor Queue Length — How many jobs are waiting to run? (Note: The CPU Load Average available on Unix boxes is a better metric since it’s time-weighted.)

Processor Utilization — On a multi-processor machine, is the work load evenly distributed? Or, is one processor pegged while the others a lightly used?

Context Switch Rate — My candidate for the single most useful “usual suspect”. Context Switch Rate offers a good measure on how “busy” your machine is — or, how much time is wasted on overhead rather than being used for application processing.

CPU Utilization — Not as useful as you might think, but can be useful if you can use it to identify what processes are running at the moment.

Memory usual suspects are:

Pages/Sec — Time spent paging is time wasted.

Pages Out/Sec (or Swaps Out/Sec) — Paging out (or swaps out) suggests desperation; active applications get swapped out because memory has become dear.

Memory Utilization — Paging is more critical, but Total Memory Utilization is useful for capacity planning and trending.

Disk usual suspects are:

Disk Queue Length — How many jobs are waiting to go to disk?

% Disk Time — How much time is spent accessing the disk

This small set of machine metrics make up a very effective group of usual suspects. Basically, you’re looking for situations where your machine is waiting for something. “Waiting” means work isn’t happening!

You can wait for CPU, you can wait for memory, and you can wait for disk access. When you’re waiting for a resource, your system overhead increases — which is reflected in spiking Context Switch rates (the reason I find it the most effective metric). It’s possible for a machine that’s only using 50 or 60 percent of the CPU to be completely hung because of system overhead.

Quite often, you’ll see symptoms ripple through these metrics when situations occurr.

For instance, suppose you have an application that’s spending a great amount of time writing to disk. This causes the disk queues to back up (because all requests can’t get through at once) and the Disk Time Percent to increase.

Because of the Disk backup, disk writes get posted to memory cache. This decreases the amount of free memory available — meaing paging increases. The system overhead involved in managing this situation means Context Switches go up.

The end result — no work gets out; everyone is waiting for something. Your CPU may not even being spiked — because you’re not waiting on CPU, you’re waiting on Disk or Memory! If you’re able to capture the processes running at this time, you can identify the jobs responsible for this situation.

Specific applications may require special monitoring — databases, daemons, numbers of users, etc. However, for general system monitoring, it’s hard to go wrong if you simply “Round up the usual suspects”!