Today’s debate is concerned with how much uptime is enough for an organization to please its customers and not over spend too much on building infrastructure. In the present scenario, customers are not content with a 24×7 uptime. They don’t only want their data retrieved on command each and every time, but they are also concerned about data representation on different platforms. To provide, such a flawless service, IT departments need to overspend on creating a superior infrastructure. More often than not IT departments are accused of overspending and still there are outages. Actually in our opinion the decision of how much uptime is necessary? Should come from the top level management. Though this question should come from the IT department itself.
What Uptime Means To Customers?
A customer who does not have knowledge of servers may never know, why it goes offline. Instead they treat an offline website like Google or Netflix as the end of the world. They are easily misguided and draw conclusions based on some information they have gathered from a source, which is unreliable. But it is not the end of the story, customers often use social media to broadcast a minor issue and publicize. Though outages are rare occurrences, but still it gets noticed by other customers, and the issue just blows out of proportion.
In all sense, an outage is catastrophic for the organization, but customer’s indulgence in discussing about such issues through social media has given a chance to the management to decide the issue of uptime and outages.
Today, if you have to measure uptime, it is done by measuring “nines”. There is a huge difference between data being available 99% of the time or 99.99%. For a normal customer it will not make any difference, but if you dig a little deeper, your perspective will change. Let’s see how these nines influence uptime?
|Uptime level Downtime per year|
|99.8% 17.56 hours|
|99.9% 8.72 hours|
|99.99% 52.58 minutes|
|99.999% 5.36 minutes|
|99.9999% 31.6 seconds|
Every company aims to have an uptime level of 99.9999%, but every nine they want to add, results in huge costs. Many companies spend thousands of dollars to achieve an uptime with six nines. Whether to add a nine or not, should be a decision, which should be taken by the top level management and not by the IT department. IT departments naturally will want to increase the uptime, but does the organization really need it is a question that should be answered by the management.
The management should understand that investing huge amount of money on servers in the hope of increasing the uptime is not an idea that may deliver desired results. The right guidance on the part of the IT department and the decision made on cost benefit analysis is the key here. The traditional way of deciding how much uptime should be changed and the decision should come from the other way around, then only this issue can be solved effectively.