Muhabirizbiz Rotating Header Image

Their

Reasons Why Webhosts are Spreading Their Customers Across Servers

As Hosting is a 24×7 business it is mainly a tough business. Everything like servers, bandwidth etc requires to be up and running 24×7. Redundancy in everything is the only solution in such a condition. One requires duplicating everything like having multiple power sources main power, UPS, generator, having multiple bandwidth providers etc. apart from this entire 100% uptime guarantee cannot be given. It is interesting to know that this kind of redundancy is wanted by web hosts from their providers. Nonetheless, hosting all their customers on a single server is the mistake made by many small hosts.

Having one high end server and filling them with as many accounts as possible is time and again considered better. Few reasons why a single server is preferred by webhosts -

1. Maintaining a single server is much easier.

2. Cost of buying multiple servers is more when compared to buying a single server.

3. Multiple O.S. Licenses and multiple control panel installations mean multiple servers. The cost of software is reduced.

4. More datacenter space is multiple servers while collocating.

5. A higher end server is considered better where customers look at the specs of a server.

6. Well-configured server when compared to poorly configured server can take more sites.

7. As the OS architecture is better a good OS such as NetBSD would take more sites.

When a single server can do the task no one would want to take multiple servers. Just selecting the right OS and configuring it by an expert is the only thing a webhost needs to do so that he/she can live a peaceful life hosting as many sites as he/she wants, perhaps even unlimited sites on a server. All this might sound too good to be true.

Most of the time even with over a 1000 sites thing looks very good. And thus it’s been decided by spammer to pay for a year and start spamming from the server. The sites start to take ages to load and the load is never below 20. All the sites will go down at one shot if the server crashes. Supposing server goes down in the morning hours of first day of the working week and the server has say about 1555 sites, then the webhost may end up with many cancellations and 755+ support tickets. Thus single servers may possibly mean:-

- As the number of support calls increases it means higher support costs.

- As the server is catering to more sites there can be higher server load. This would result in slower sites.

- As the number of users per server is higher there are more risks. More users per server is difficult to maintain.

Total throughput of the Ethernet card is another important aspect that many hosts forget when hosting on a single server. At any instant of time a single card would always have its limitations on amount of data it can push. If many of these sites are accessed at the same time, the Ethernet card would fail to push the data quickly when multiple sites are being hosted.

Lastly, lower-end servers consume less power than high-end servers. They are also heated up more than regular servers. This over heating leads to faster hardware failure.

On a typical server “How many sites can a webhost host?” No exact number is there for this. Just how many sites per server are ideal has to be determined by every host. Factors required to be considered to reach this figure: –

- The ratio of static to dynamic html sites on the server.

- Type of O.S. and software used.

- Different packages being offered by webhosts. Higher bandwidth, more mailing lists, databases etc. would mean higher server loads.

- Maintenance of servers. Servers that are not maintained on a regular basis have lesser sites hosted.

- Place where customers are located, as a lot of sites from the same time zone would have higher usage at the same time.

The load ought to generally be lesser than 1 for a single processor server. One can see the CPU load by using the w or top command on Linux. Or merely cat /proc/loadavg . 0.5 means that only half of the CPU time is utilized whereas the load average of 1 means the entire CPU cycle is being utilized by processes. Experiments determine the right load average.

A lot of importance should be given by webhosts to the amount of RAM as well as the type of network card used by the server. While taking server one should always go for high-end Ethernet cards. The maximum throughput of data would be ensured.

Web hosts cannot replicate the setting on servers because of the demand for different O.S. and control panels. Maintaining the servers will not be a problem if a good policy is in place. Webhost are allowed to broaden their offerings to the end users in terms of the O.S. and control panels offered by multiple servers.

Only when a server crashes or when a server is hacked the biggest benefit of having multiple servers is being understood. The number of trouble tickets is reduced in such cases, as every site that is hosted by the webhost is not affected. Instead of having an angry crowd of 1555+ customers at one shot it is better to have 250 customer sites down at a time in case of hardware failure. Multiple servers also mean that the webhost can spread the name servers across the servers.

Webhosting can take a lot of your attention unless your webhost gives you good uptime and prompt support. Hostgator is one of the best web hosts in the hosting coupons lists where you can host your website and blogs and also get upto 20% discount

Self Insuring Property Owners Become Masters of Their Own Destiny

In the aftermath of Hurricane Katrina, insurance premiums for property owners in vulnerable coastal regions skyrocketed. The Grand Strand was no exception, even though we have not had a major hurricane since Hurricane Hugo in 1989. Unfortunately, insurers were backed into a corner. Massive losses on Katrina claims changed the risk tolerance of many underwriters. At the same time, rising property prices during the inflating of the real estate bubble necessitated increased premiums to cover more highly valued properties. As many insurers cancelled policies in coastal regions, even those not impacted by Katrina, coverage became much scarcer, further driving up the costs to insure property. Ultimately, property owners throughout the Southeast would bear the burden of this.

Insurance is a calculation of probabilities and risk. What is the probability that an oceanfront property will experience a loss in a particular year? Furthermore, what is the projected magnitude of the loss (partial loss vs. total loss)? These are some of the factors that drive insurance premiums. For example, the risk of total loss of an oceanfront home during a hurricane is far greater than the risk of a total loss of a home ten miles inland. Consequently, the risk of a total loss of an oceanfront home is far greater than the risk of a total loss of an oceanfront condo tower constructed of concrete and steel. Underwriters assess these risks against the replacement value of the property to derive the policy premium and coverage limits.

Following Katrina, many property owners experienced an increased premium for an equal (or in some cases a lesser) amount of coverage with significantly higher deductibles. As many insurers cancelled policies and pulled out of coastal regions altogether, property owners were forced to accept higher costs to insure the property. Though many property owners would have liked to drop coverage altogether, assessing the risk of loss a lower probability event than warranted by exorbitant premiums, the holders of mortgages on the property mandated adequate insurance coverage. Homeowners, then, were trapped in a catch-22.

As the costs spiraled out of control, property owners along the South Carolina coast encouraged legislators to take action. This action came in the form of an expansion of the wind pool. There were high hopes this would drive insurance premiums down. Surprising to most, the impact of the wind pool expansion has only been minimal. Property owners now want an elected insurance commissioner as opposed to an appointee. But will this solve the problem?

It is unlikely that this will solve the problem or that insurance premiums for property owners in coastal regions will ever be inexpensive. Over time in the absence of any catastrophic events like Katrina, more underwriters may offer coverage to property owners in these vulnerable areas. This additional competition could drive premiums down to more affordable levels and provide property owners with more favourable policies. But another major storm that results in heavy losses would start the vicious cycle of higher costs all over again.

Property owners, it seems, are left with little choice but to bite the bullet and accept higher insurance costs as a way of life in coastal areas or to take matters into their own hands by self insuring their properties as a viable alternative to traditional insurance.

Self insuring, however, is expensive for property owners to initially establish, and many lenders may not accept self insurance on properties. Typically, self insuring involves creating a trust into which funds are placed periodically to help build reserves to cover potential losses. The trust may then buy reinsurance for low probability events that result in extreme losses. Reinsurance is generally less expensive given a high deductible. Thus, self insuring is likely too costly for individual homeowners. However, this may be more feasible for property owners in condo towers where self insuring costs would be widely spread amongst the owners.

From a financial perspective, the cost to insure a $100 million condo tower is more than the cost to insure a $2 million oceanfront home or a $300,000 home ten miles inland. But the oceanfront condo towers, typically constructed of concrete and steel, are built to a much higher standard than many homes. Many oceanfront condo towers, even during Katrina, withstood tremendous damage but were not total losses—the structures and their integrity were not compromised. Most damage was cosmetic—broken windows, roofing, facades, etc or water damage. Whilst insurance would cover some of this damage, property owners are still on the hook for the deductible. With self insuring, property owners’ advisors assess the expected amount of loss in the event of a disaster. Contributions may then be made annually, quarterly, or monthly or at one time through a special assessment. In the event of a claim, the funds from the trust are used to cover repairs, etc. In the event of an extreme loss, the reinsurance kicks in after the trust pays the deductible.

There are pros and cons to self insuring. Obviously, if the self insurance trust does not have sufficient funds in the event of a claim, the property owners must come up with the difference. The benefit is that the trust is an asset of the homeowners. The payments are not simply paid to the insurer and gone forever. The trust is held by trustees as an asset of the property regime and can be used for smaller claims that would typically be under the amount of the deductible on a normal insurance policy. Over time, the trust’s assets grow, particularly in the absence of any claims, thus raising the possibility that the property owners may not have to make additional contributions at times.

Whilst this could work for smaller homeowners, it is clear such a plan would be more costly for the individual. However, over time, as underwriters do not have the exposure to higher dollar value condo developments, costs to individual homeowners in the same area may fall as the underwriters’ concentration of risk in one area falls.

It is clear that climbing insurance premiums are a burden on property owners and a drag on real estate markets. Whilst many proposals have been made, none seem as effective as self insuring. Such self insuring regimes allow property owners to control their own destinies rather than be at the mercy of insurers.

Robert M. Clinger III & Sebastian G. Perey

Copyright 2008 Thinking Outside the Boxe

http://www.ThinkingOutsideTheBoxe.com