These 4 Failures Cause The Majority Of Website Crashes

According to Venture Beat, downtime costs the average small business $12,500 per hour. This figure means that if you create a business website that’s up 99.5 percent of the time, it’s down 44 hours per year at a cost of $549,000 per year. When you create a website, make sure your hosting provider can prevent significant downtime. There’s no such thing as 100-percent uptime, but there are ways to defend against the top four causes of website crashes.

1. Power Failure

Power outages are the top culprits behind website crashes. Disasters like fires, tornadoes, hurricanes, and winter storms are big causes of power outages. No hosting providers can hold off a hurricane, but they can put protections in place to keep your website up and running.

Reliable power sources. In addition to having reliable utilities providers, your host should have uninterruptible power supplies (UPIs) for servers, which keep them up and running for a short time until onsite backup generators start running. Also, generators should be maintained and inspected regularly so they don’t fail when the power goes out.

Usage monitoring. Overuse can overload circuits, causing local and sometimes widespread power outages. Your hosting provider should monitor machine usage and have alerts set up to warn them about overloaded circuits.

Plans for temperature control. When the power to a datacenter’s cooling system goes out, the heat produced by machines can cause equipment failure. Your host’s datacenter should connect its cooling system to UPIs, and it should keep an extra chilled-water storage tank on-hand to buy the datacenter some time until the generator starts. Also, its cooling system should be designed to handle larger than normal capacity so that the temperature comes down quickly once power is restored.

2. Hardware Failure

Hardware failures have caused several public website crashes over the past few years. In 2013, Yahoo Mail’s free service crashed for four days because of what CEO Marissa Mayer called a “particularly rare” hardware outage on storage systems. In that same year, routine server failure at a Utah datacenter caused outages for multiple hosting providers.

Hardware failure can happen during maintenance, periods of heavy traffic, or unexpected events, such as a mouse chewing through a server’s power cable. Your hosting provider has to prevent single points of hardware failure from shutting down your business website. In addition, your hosting provider should be conducting drills so that the IT team knows how to respond in case of hardware failure. Your provider should also regularly backup your files, preferably onto physical machines.

3. Software Failure

Software failures that crash your website can happen for different reasons. Sometimes, datacenters operate using an old operating system (OS) as a cost-saving measure, and the old OS gives out. Also, untested patches sent from software providers can corrupt applications, making systems unavailable. Another cause of software failure includes viruses and other types of malware. When you build a website or update your existing one, make sure you prevent these failures by checking that your hosting provider’s datacenter uses up-to-date operating systems from trusted providers, and ask what security measures your hosting provider takes to protect your applications and your data. 

4. Human Error

In 2013, the NASDAQ exchange went down for almost an hour. The outage caused the NASDAQ composite index to freeze, and it caused trading to stop on certain options related to indexes. So what brought the NASDAQ to its knees? Simple human error. “The disruption was caused by a human error performing an operational function which resulted in the incorrect delivery of data to the index distribution system,” NASDAQ said in a press release about the outage.

No amount of equipment redundancy, backup power, network security, or ingenious datacenter design can prevent human error. For this reason, it’s important to consider how the datacenter expects its workers to function. When you choose a hosting provider, ask about how well the human operators follow procedures, how the datacenter ensures physical security, and how the datacenter treats its workers.

When you’re choosing a hosting provider, don’t let fancy website creation distract you from the most important issue: whether or not the host keeps websites up and running. Ensure that uptime guarantees are in your service level agreement (SLA), and ask for historic uptime performance information. A website is no good to you if it’s not available when you need it.