What are the Nines of Availability?
What does it mean and what does it imply that a company needs an availability requirement of 99.78%? We'll tell you...
In the previous entry of this blog, we defined the concept of availability and also learned a simple way to determine the necessary level of availability. This percentage-based result determines the technical and economic implications of the solution to be implemented in our factory.
Returning to the case of the previous entry: a company in the pharmaceutical sector whose objective is not to lose traceability data and whose availability requirement is 99.78%. The following question arises: what does this percentage mean and what does it imply?
Basically, it implies that according to our availability objective, the system must be available to users 99.78% of the time, that is, we can allow ourselves 86 minutes of downtime per month or, what is the same, 17 hours of downtime per year. According to our previous example, it could mean that we can afford to lose data for these periods of time, without major economic or corporate image implications.
High availability and fault tolerance:
In the following table, we will see the translation of percentages and times:
A system that exceeds an availability percentage of 99.5% is called a high availability system. And a system that guarantees an availability level of 99.999% or 5 nines is because it is a fault-tolerant system.
What is the cost to the company of a system failure?
The previous table also allows us to determine, based on our cost of inactivity per hour, the total annual cost for the company if the system stops working.
For example, if for this pharmaceutical company the cost of losing production traceability data due to fines and penalties is €10,000/hour, look at the annual cost according to the level of availability: 
With this data, what availability percentages would you require for this particular system?
In our next entry, we will talk about availability, high availability, service levels, continuous availability, fault tolerance, and disaster recovery strategies.
