Deciding what level of availability you need

After understanding availability at a basic level, it is important to assess your individual availability needs. Higher availability is more costly than a lower-level availability. You must balance your needs and services with the overall cost of implementing and maintaining these availability solutions.

You want to be sure that you have analyzed your business needs thoroughly in order to decide what level of availability you can afford to maintain. To decide what level of availability you need, consider the following questions:

Do you have any applications that require 100% availability?

In most cases, you can achieve a high level of availability by implementing sound processes and systems management practices. The closer you need to be to continuous availability, the more of an investment you must make. Before you make that kind of investment, you should be sure that you require that level of availability. The following figure shows how different techniques can improve availability, but can increase the price you must pay for it.

Determine your level of availability

If your requirements for levels of availability increase, you can consider multiple system availability solutions, such as clusters.

How much downtime is acceptable to you?

It is helpful to know the amount of downtime that each level of availability represents. The following table shows the amount of downtime expected for different levels of availability.

Level of availability Downtime per year
90% 36.5 days
95% 18.25 days
99% 3.65 days
99.9% 8.76 hours
99.99% 50 minutes
99.999% 5 minutes

Along with knowing how much downtime is acceptable to you, you need to consider how that downtime might occur. For example, you might think that 99% availability is acceptable if the downtime is a series of shorter outages that are distributed over the course of a year, but you might think differently about 99% availability if the downtime is actually a single outage that lasts three days.

You also need to consider when a downtime is acceptable and when it is not. For example, your average annual downtime goal per year might be nine hours. If that downtime were to occur during critical business hours, it might have an adverse effect on the bottom line revenue for your company.

What level of access do your customers need to your business?

It used to be that customers and business partners accessed your business from 9 a.m. to 5 p.m., so it was realistic to expect that your system only had to be available during those hours. However, the Internet and a diverse global marketplace have changed that expectation; customers and business associates might expect to have access to your company's data at any time of the day or night. Your working hours might be hours or even days different from your global business partner or customer. You must determine what your customer expectations are, and what is realistic with regard to those expectations, as you determine what level of availability you will maintain.