When talking to customers and prospects I like to ask, “What is the one thing HOSTING should do really, really well to succeed?” (I was heavily influenced by a cowhand named Curly in City Slickers).
I was surprised by how clearly the answer was expressed in a survey at our annual customer summit. About 70% of the votes, in different words but the same basic concept, came back saying “100% uptime.”
At HOSTING we believe that every business critical application deserves a highly available (i.e., fully redundant) architecture with disaster recovery built in. There are, of course, different solutions that deliver different performance and price points, but for example, we provide a storage architecture built with multiple storage processors so that physical infrastructure could fail without impacting the customer. Then, from a capacity management perspective, we build our cloud to have less than 100% of one storage processor utilized so that customers would not even see a degradation of performance in time of physical failure.
There are also 3rd party software technologies out there such as VMWare’s vSphere that utilize technologies such as vMotion, DRS, and Storage vMotion to ensure that, as customers scale their environments through our Customer Portal, they will have those resources available to them without service interruption.
The last example I’ll give – and it’s not even close to an exhaustive list – is that our DNS Architecture includes a cluster of caching servers to not only enable quick response time, but equally as important, to allow for physical failure without creating an impact to DNS availability.
On the surface – and if you’re only thinking about technology – this is a good start. But availability requires more than just technology. Consider these three basic functions:
- Capacity Management – making sure you have enough “stuff” to meet demand and growth
- Proactive Maintenance – making sure we are maintaining the “stuff” so it works as designed
- Monitoring – making sure you are always watching and reacting when “stuff” is about to break or does break
The Product teams work to understand the patterns of business activities that exist and what customers require to solve their problems. All of this comes down to a service being fit for use – not only for availability; they must consider performance, feature and function, as well.
Engineering makes sure that the solution is built to meet the performance demands for the patterns of business activities for which the service is designed, and that it is fit for use.
Operations then manages all three aspects on a 24/7 basis and “tunes the dials” for customers based on their specific needs.
To think of it another way, customers need a specific SLA: Product teams identify the SLAs, Engineering designs the solution to meet the SLA, and Op’s manages the solution to meet and exceed the SLA.
All of this is then wrapped in a big blanket of Continuous Service Improvement. As time marches on, the patterns of business activities change, as do the needs of the customer. What is fit for use today will not be tomorrow. It takes all three groups – Op’s, Engineering, and Product – to identify the future customer problems so that HOSTING services that are working today can evolve and continue to solve problems for customers in the future.
The bottom line is that, like it or not, electrical and mechanical equipment fails over the long-run. Your business critical application should have the right architecture, built on best-of-breed technology components, managed by a team that specializes in building and operating high-end product applications. This three-pronged approach is much more fool-proof in terms of failure.
This is why, at HOSTING, we are proud of the fact that we have over 200 cloud certified engineers, working in dedicated customer teams, within an ITIL framework. We take our 100% uptime mission seriously!