Importance of Availability

Availability becomes increasingly important as businesses continue to increase their reliance on information technology. As such, the availability of mission-critical information systems is often tied directly to business performance or revenue. Based on a system's given role in an enterprise, downtime can lead to other negative consequences such as loss of life, customer dissatisfaction, loss of productivity, bad press, or an inability to meet regulatory requirements.

Table 1: Average Cost of Unplanned Downtime for Various Industries

 Source: Contingency Planning Research, 1996. © Eagle Rock Alliance, LTD. All Rights Reserved. For more information about Eagle Rock Alliance, see http://www.eaglerockalliance.com/.

There is a common misunderstanding of the terms used in our "tech community." 

• Fault Tolerance is NOT a RAID configuration - Fault Tolerant means tolerant of faults - a fault is a error or malfunction that can cause a server or any tech equipment to crash 
• High Availability is no where close to total availability - Clusters have over 8 hours of downtime as an average. Clusters are not a solution for 100% business continuity and availability. They can help secure and protect data - but really that is the only advantage clusters have.
• A fault tolerant server is a server that can give 100% availability because it is 100% redundant - that is until servers are engineered or made so well that they just can't crash - until that day comes - there has to be a technology that offers 100% redundancy. Please look table #2.
• All servers crash or need updates - even Linux servers - downtime IS DOWNTIME - too many times have I heard techs stated that their server never crashes or has downtime. It's wishfull thinking but a joke to think otherwise. 

Table 2

However, not all downtime is equally costly; the greatest expense is caused by unplanned downtime - which the NEC Fault Tolerant server elimanates. Outside of a system's core service hours, its amount of downtime-and corresponding overall availability level-may have little to no impact on a business. In contrast, if a system crashes during core service hours, the result can be significant financial impact. Because unplanned downtime is rarely predictable and can occur at any time, companies looking to minimize risk should evaluate the cost of unplanned downtime during core service hours.

For example, in projecting the consequences of planned versus unplanned downtime, consider a modest manufacturing scenario where a system is used only during plant hours. Within these intervals, each hour of unplanned downtime costs the company an average of $28,000. However, on evenings and weekends, the system can be taken offline for maintenance or application upgrades with no impact to the company's operations. Therefore, while the system's overall availability level may be only two nines, the corresponding 87.6 hours of annual downtime will not have a financial impact, other than possibly paying an IT staffer to work the occasional evening or weekend.

On the other hand, systems supporting functions such as telecommunications switching or a city's 911 emergency police, fire, and ambulance dispatch operations require full-time, round-the-clock availability-often called continuous availability. In these situations, there are no off-hours; every second a system is down leads to an interruption in phone service for thousands of people, or, even more damaging, delayed response to a life-threatening situation. System administrators in these scenarios face an additional challenge: how to perform periodic hardware maintenance or install software upgrades without compromising availability.

Fault-Tolerant Servers

Fault Tolerance Defined

As operating systems, applications, drivers, and other software-based solution components become more reliable, hardware-related issues and failures play a larger relative role in determining a solution's total downtime. One approach to minimizing these causes of downtime is through the use of fault-tolerant servers, combined with software that supports them.

Put simply, fault-tolerant servers are those that have complete redundancy across all hardware components. If a primary component fails, the secondary component takes over in a process that is seamless to the application running on the server. As such, fault-tolerant systems "operate through" a component failure without loss of data or application state. This behavior is different from software-based failover clustering, in which a hardware or software failure on one server causes the workload to be shifted to a second server.

Although a system may have some redundant components, it is not necessarily a fault-tolerant server. Most high-end servers employ at least some redundant components to eliminate common points of failure (for example, hot-swappable power supplies, ECC memory, or multipath I/O adapters), but they will still fail when a non-redundant component such as a microprocessor or memory controller fails. True fault-tolerant servers, however, employ complete redundancy across all system components, ensuring that no single point of failure can compromise system availability. Some fault-tolerant servers extend this level of redundancy across data center boundaries by allowing the server's redundant subsystems to be installed in separate yet connected locations.

Note: Servers with only selected redundant components cannot deliver the same level of hardware reliability as a fully fault-tolerant system, but they do offer greater reliability than servers without redundant components. As such, they may present a cost-effective way to decrease unplanned downtime in situations where full fault-tolerance is not economically justified.

Superior Return on Investment

Because of the benefits, fault-tolerant solutions on Windows typically carry a far lower total cost of ownership than solutions built on other fault-tolerant platforms. Companies switching from proprietary fault-tolerant solutions to fault tolerance on Windows can reduce costs without compromising availability. Why should your company look into a Fault Tolerant server? What are the issues that plague your company? Maybe you are looking at increasing employee production, reduce the hardware and software costs related to clusters, eliminate planned or unplanned downtime or just the simplicity of our Fault Tolerant servers - Similarly, companies in industries with lower costs of downtime-such as manufacturing-can now improve the availability of mission-critical systems and still realize a positive return on investment in a reasonable time-frame.

For more information - Contact Us!