Imagine this scenario–you are sitting in a typical meeting room. At the table there are a few representatives from your business sitting across from you with arms folded and a serious look. The topic of service availability has come up. When you begin the discussion with the business around expectations, they reply, “Well, frankly we need this service running all the time.”
This is the response you had prepared for; you take out your pen and then write down a number on a piece of paper which indicates the annual cost of continuous availability, fold it and slowly slide it across the table in their direction. The person in the middle looks at their colleagues on either side and picks up the paper. They pause.
They look at you for a moment before they unfold the paper. Once they open it up you can see that there is a slight gasp, their eyebrows rise, and there is a brief hesitation… Then they reply,
“8 to 5 Monday through Friday will be just fine.”
While discussions like this are the things of satire, there is an ounce of truth, in that the customers just want to have their services up all the time. The key question is how to optimize availability versus cost of having that type of availability. I spoke a bit about this in my last article.
How do we go about this effectively, one might ask. As always, a good place to start would be to determine what the business really needs regarding service availability:
- Business hours
- 24 x 7 x 365
- Can we allot for maintenance outages etc.
Once we know what the business needs from an overall service availability standpoint, we will need to understand what risks are present from an unavailability standpoint. Depending on the capability of your other service management processes, you may be in a better position to identify:
- Volume of incidents which impact the service
- Monitoring the components within the service or the service as a whole via Event Management
- The level of architectural redundancy built into the service and if there are any single points of failure or weaknesses
There are several ways to identify what can contribute to the unavailability of a service; whichever you choose at the end of the day should allow you to strategize on how to mitigate those identified risks. Changes may be required at an infrastructure level if we have single points of failure. There may also need to be changes in the way which IT provides support during incident outages. These along with other factors you discover during the assessment of the service will identify the costs that are associated with the service unavailability.
This price tag will be used to balance what levels of design and support are required for the availability of service we are discussing. The fact of the matter is that nothing is unbreakable; establishing the availability requirements to manage failures for your customers where they make sense is paramount. This in turn will translate to value for your customer.
For more brilliant insights, check out Ryan’s blog: Service Management Journey