I have been concentrating quite a bit of time lately on the subjects of Availability and Uptime: how they are measured, what they mean and how to make use of these meters.
I've been renovating my home and my two most vital tools are a measuring tape and a square. The tape allows me to measure 25 foot lengths to an accuracy of 1/16th of an inch and the square ensures that my lines and cuts are straight and properly aligned. Those two tools have specific attributes and functions. They are both tools used to measure to some degree of accuracy but they are not interchangeable. I can use a tape measure to sort of measure out a line or cut that will be ‘square’ and I can use a square to sort of measure the size of a room but I don’t. I use the right tool for the right job - the square to make a square cut and the tape measure for determining lengths. When it comes to assessing performance, it is equally important to use the correct metric and method. Two terms that are similar, but each of whom have specific attributes and functions are Uptime and Availability. Let’s discuss their uses, where they are similar and where they are different.
Good old Wikipedia tells us that Uptime is the measure of the time a machine or system has been up without any downtime. Availability is a measure of the time that a machine or system is capable of being used. Both of these terms involve a measurement of time relative to some degree of functionality, but they are not the same because they do not measure the same things. When placed in the context of measuring a service, Uptime, refers to a user's ability to utilize that service. If that user cannot use that service, then from their perspective, that service is DOWN, regardless of the cause. Availability, however, is a measurement of whether that service is available to be used. Meaning that it may be possible that some users are able to utilize the system while others cannot. I know, it sounds convoluted, but the distinction is important. In a product suite such as there are a multitude of different components all working together to provide a product – hosted contact center services. Telecom networks, data networks, computers, web servers, databases and applications must all work in conjunction for every user to be able to utilize every service at any given moment. That said, it is possible for any of those components to have problems that impact one set of users and not another. Suppose the loss of a data connection temporarily interrupts the data connections of three user locations. Those three locations may be temporarily unable to communicate with the platform. Yet, the platform itself, and the routers, the switches, the servers and the software are working, and users not on that failed data network are completely functional and able to use the services. In that scenario, the platform would be considered to be 100% Available. The Availability on the data network, well, that is another story.
Hence, we see that as a measure of performance, both Uptime and Availability have their specific use and meaning and each needs to be used in the appropriate fashion. In my next blog, I will continue to discuss Availability, its use as a metric and methods to quantify its measurement.