Do you want to be like Amazon?
- Summary:
- Leading cloud services providers such as Amazon and Salesforce.com have made their ability to execute operational efficiencies well in excess of most end user organisations a major selling point. Can you do the same?
Have you got cloud envy?
Leading cloud services providers such as Amazon and Salesforce.com have made their ability to execute operational efficiencies well in excess of most end user organisations a major selling point.
But according to a new report from Forrester Research, it is possible for end users to learn from the data center and infrastructure lessons of the cloud experts.
The report by James Staten and Richard Fichera - Five Data Center And IT Infrastructure Lessons from the Cloud Giants - is based on interviews with executives from 10 of the top cloud and hosting providers, including Amazon, Colt, Microsoft, Rackspace Salesforce.com and Saavis.
It notes that organisations can end up with an inferiority complex when they compare their own operational strategies to those of the cloud industry:
The economics and efficiencies of the world’s largest cloud and hosting service providers are an aspiration for some in enterprise IT — but a bludgeon for others.
If you fall into the latter category, the capabilities of these digital giants seem like a nagging reminder that you will never match their efficiencies and should just fold up your data centers and dump all your workloads into the “cloud.”
The report admits:
But it adds:Many of the cloud and hosting provider accomplishments, especially on infrastructure cost, can only be realized when you purchase thousands of servers at once in uniform configuration.
While you do not have their scale of purchasing power, IT infrastructure and operations (I&O) professionals can benefit by adopting their approach to infrastructure architecture and operational process.
The major difference between large enterprises and cloud services companies is how they approach operations. The degree to which these web properties automate processes at all levels remains a formidable barrier to enterprises ever achieving the same levels of opex efficiency.
However, as the solutions available to enterprises in the loose and constantly changing categories of “private cloud” and “service automation” continue to evolve, the gap will narrow.
Staten notes:
There are several key differences between you and the cloud leaders, many of which are detailed in the table below. Perhaps the starkest however is that for the clouds, they are the product. And that means they get budgetary priority and R&D attention that I&O leaders in the enterprise can only dream about.
But at the end of the day, clouds are a bunch of virtualized servers and storage racked and stacked in an air-conditioned building, just like you are.
Five things to do
Forrester highlights five key lessons to be learned:
Bring IT process automation to the facilities Level
Forrester notes that cloud companies have to contend with massive scale and density, but don't have to deal with the heterogeneity of load and applications that afflict the typical enterprise data center.
It advises:
Many of the advanced power and cooling management techniques used by the large web properties have become readily available to modest-sized enterprises through the various data center infrastructure management (DCIM) products that have been brought to market over the last few years.
These solutions deliver sophisticated management of important opex KPIs, such as PUE, actual and forecast power and cooling data at multiple granularities, and UPS and battery status, that approach the sophistication of the larger properties.
Prioritize speed-to-market and standardization over customization
Forrester argues that not only are customized infrastructures unsustainable, they are often not even what the business wants.
It advises:
Leverage virtualization abstraction and workload optimization via pools of consistent applications to drive true standard infrastructure implementations. Only through these efforts can you truly deliver the cost efficiencies, strong performance, and availability these services demand.
Automate basic IT infrastructure processes
Forrester suggests that the limiting factor in automation initiatives at most enterprises is the inability of most I&O groups to develop processes that lend themselves to automation, and then define these processes sufficiently well that they can be then automated.
The research firm notes:
Automation tools are maturing and getting more integrated. While standalone tools still dominate, suites are improving quickly, driven mostly by a desire for more complete cloud service orchestration. Tools remain in need of much more development, but key task execution and process flow capabilities are now solidly in place.
Shift from infrastructure management to infrastructure service delivery
Few applications today need unique configurations or truly demand specific hardware dependencies, states Forrester, meaning that organisations need to start shifting workloads to a small set of standard configurations that will be automatically deployed.
It argues:
- Think like a service provider and embrace customer service.
- Shift your administrators to service managers.
- Mask infrastructure complexity with service-level capabilities.
- Remember to give administrators a career path in the service delivery world.
Break down organizational silos
To be successful at implementing the process changes, Forrester urges organisations to work across silos to ensure consistent operations and engage AD&D and EA and shift to modern application architectures.
Forrester argues:
As cloud services proliferate, application architectures are beginning to shift to new models where availability, performance, and other needs are fulfilled at the software layer, not with infrastructure dependencies.
Embrace and encourage this transition as it empowers your shift to DevOps, driven by infrastructure abstraction, standardization, and efficiency.
Verdict
Sound advice.
Disclosure: At time of writing, Salesforce.com is a premium partner of diginomica.