Cloud computing is here, and it’s just the tip of the iceberg.

So says Manon Buettner, Principal of Nuvalo, a Bellevue, Wash.-based consulting and brokerage firm specializing in simplifying cloud migration for companies, who notes the rapid deployment of enterprise use as a clear indicator the cloud is here to stay.

“Cloud use is now a legitimate line item for companies, who realize it’s become not only a necessity to store increasing amounts of data in this fashion, but it also provides a terrific ROI when considering infrastructure and other associated costs with maintaining such data in-house these days.”

In fact, the latest Forrester Research survey indicates that nearly half of all enterprises in North America and Europe will set aside budget for private cloud investments in 2013 – and nearly as many software development managers are planning to deploy applications to the cloud.

“That’s a lot of deployment,” adds Buettner.

So while 2013 looks to be The Year Of The Cloud, Buettner is encouraged that companies will now start incorporating cloud services and platforms into formal IT portfolios.

But with this movement also comes the usual growing pains, and Buettner offers 12 predictions on cloud computing in 2013 – both good and bad – based on her own observations, in conjunction with thoughts from other leading industry experts:

1 – Not Everything Will Go To The Cloud

Gone are the days of “through everything on the cloud . . . just because.”  We now know what makes cloud platforms different from traditional virtual infrastructures and traditional hosting environments.  This enables us to make more educated decisions on what makes sense to move, and what makes sense to stay.

2 – Cloud & Mobile . . . One In The Same

We’re finding more mobile applications connected to cloud-based back-end services that are able to respond to mobile client engagements and shield data centers from such traffic.  In fact, nearly every SaaS application has a mobile client now.

3 – Stress-Free Cloud SLAs

The optimal practice for cloud application design and configuration is to build resiliency into the application – rather than expecting it from a cloud platform.  This enables achievement of any type of service-level agreement regardless of the base SLA provided by the cloud platform.

4 – Cost Modeling Improves

If you’re seeking the best ROI from the use of cloud services and platforms, you need to actively model the cost profile of your applications, monitor resource use, and adjust accordingly.  Many cloud cost-monitoring tools (Cloudyn, CloudCruiser, Cloudability, et al) now exist to quantify such a move in dollars and cents (or is it that “dollars and sense?”).  Good cost management in 2013 will be an absolute must.

5 – The Cloud Will Improve Backup & Disaster Recovery

Instead of enterprises buying resources in case of a disaster, cloud computing and its pay-per-use pricing model will let you pay for long-term data storage while only paying for servers when testing or declaring a disaster.  The cloud is turning the cost of storage upside down faster every month, and what was cheaper to back up to traditional DR storage last year will be cheaper and easier to put in the cloud is short order — and faster to recover.

6 – The Cloud Will No Longer Mean Commodity

We’re already seeing cloud services backed by high-end hardware, offering GPUs, SSDs, and other clearly non-commodity infrastructure options.  In 2013, expect to see the proliferation of these types of choices as cloud providers leverage them to meet specific market demands and to differentiate competitively.

7 – The Cloud Will No Longer Mean Just AWS

Amazon Web Services may own up to 70% of the cloud platform market share, but that will change, as Microsoft and Google, among others, have made substantive strides in grabbing their piece of the pie.  And more public cloud platforms will no doubt unveil themselves in ’13.

8 – Advanced Virtualization, I’ve Met A Cloud . . . And You’re No Cloud

Many enterprise I&O departments renamed last year’s VMware environment a private cloud so they could say “mission accomplishment” to the brass.  Very few of these environments offered self-service to the developer, fully-automated provisioning, standardized services, or cost transparency.  Quite simply, it’s not the cloud.  And more will see this in 2013.

9 – Development, Shmevelopment . . . And The Cloud Is No Different

The majority of languages, frameworks, and development methodologies used in the enterprise are also used in the cloud.  Now, what’s different isn’t the coding, but the services orientation and the need to configure the application to provide its own availability and performance.  Some of the best practices and cloud services may be new, there are few excuses for a well-trained developer to not be productive in the cloud.

10 – The Cloud Will No Longer Be An Option, But A Requirement

Currently, more than eight in 10 companies use some form of cloud technology.  Simply put, cloud adoption is gaining traction, thus making it a legitimate necessity these days among enterprise businesses.

11 – The Great Cloud Debate Continues . . . Public, Private Or Hybrid?

But that’s OK, as each can be designed for a different purpose, thus allowing some flexibility in the design of the implemented cloud technology.  Companies that can afford it generally opt to create a customized private cloud and/or use a hybrid option where the private cloud stores the data and the public cloud provides the functionality and collaboration capabilities.  But, as private cloud options become more affordable and organizations see the value in this type of system, adoption for this type of cloud in lieu of the hybrid approach may take over.

12 – IT Departments, They Are A Changin’

This maybe be obvious, but as more info is stored in the cloud, the IT department will have to adjust their skills by learning the ins and outs of the new technology and serving as a resource on how to support the technology.  Cybersecurity training will become more prevalent, for example.

So . . . those who have already adopted cloud technology strategies have found that the flexibility of cloud technology provides a cost-effective solution to increasing storage space, as their big structured and unstructured data needs a place to rest.  Storage needs are increasing, and as Buettner looks into the crystal ball, she also wonders “how much more data will we all have a year from now?”

Data centers are on the move. Countless companies across all business sectors are relocating their computer facilities in search of more efficient, effective and secure IT operations.

In fact, according to a recent Gartner Group study, more than 70% of the Global 1000 organizations will have to either move or modify their data center facilities significantly during the next five years.

“Data center relocation is due to a number of factors, the most common being the need to update aging infrastructure,” noted Manon Buettner, Principal of Nuvalo, a Bellevue, Wash.-based consulting and brokerage firm specializing in simplifying cloud migration for companies.  “Aging data centers can lead to a host of problems, but a poorly planned and executed move can wreak even more havoc.”

And if the move is not a success, it can disrupt business operations for days or even weeks.  Thus, Buettner offers five basic ways to most effectively plan and manage a data center move:

1 – To Build Or Lease?

When a company owns its own data center, it assumes greater maintenance costs, security expenses and other risks. Building a new data center from scratch requires a large capital outlay.

And although building a new center gives an organization maximum flexibility and control, for most companies this option is too far removed from their core competencies to be realistically considered.

But leasing also presents a host of challenges.  Landlords can raise the rent or make other demands.  So when leasing, long-term lease agreements are an absolute necessity.  Relocation is simply too expensive, time consuming, and risky to be forced prematurely.  Contracts should include extension provisions. It also means leases should contain expansion

options for future growth and possibly purchase clauses should the landlord falter financially and be forced to sell.

Whether a company decides to build or lease, it cannot afford to trust to luck (or to any provider’s unvetted expertise) that the design of the data center will fulfill its current and future needs.  And whether it builds or leases, a company can (and most often should) outsource the design work if it lacks internal expertise in the design of state-of-the-art high-density data centers.

If a company is leasing a new data center, it is of course constrained in design by the available space and future expansion rights. This will make optimizing the available space even more critical.

2 – Site Suitability Analysis

This must be conducted prior to either leasing or building, and include factors such as:

• Geography – Must be located far from high-risk sights due to Mother Nature, as well as near major highways and flight corridors, on high ground and protected.
• Communications – Must have multiple, fully diverse fiber connections to network service providers.
• Ample Power – Must have easy access to abundant power from multiple sources of electricity, taking advantage of low cost providers whenever possible.
• Space – Analyze planned business growth before moving.  Make sure you plan for additional space and power you’ll need 5-10 years down the road.

3 – Plan, Plan, Plan . . . And In Advance, Advance, Advance

It’s always recommended to bring in an outside expert to help plan your move.  This is too big of a project not to, as a sub-par move can be disastrous.  Buettner recommends these basic steps:

• Determine Maximum Allowable Down Time Acceptable For Move – With proper planning, a move can be accomplished in a time window that is appropriate for that particular company – from zero downtown, to several hours or even a day or two. Determine what’s best for your operation in advance, and plan with that window in mind.
• Identify Server, Storage & Application Dependencies – As enterprise applications
• cross departmental boundaries, it is important to define functional “move bundles” to assure that full business process support is returned to service as quickly as possible.
• Review All Relevant Agreements & Contracts – This should include vendor leases, maintenance agreements, warranties and insurance policies. Be prepared to determine how requirements or restrictions may impact the move’s success or introduce risk.
• Map Out Move Resources – Who will be involved and what will be their role be prior to, during and following the move? What supplier and service contractors will be involved?  Make sure the chain of command and responsibilities are clear.

Also create a “runbook,” which is similar to what NASA creates for launches – complete with a thorough, minute-by-minute, plan for your move events.  You can never be too thorough, and this runbook should encompass everything from application shutdown to system backups to physical move activities.

4 – Minimize Change & Complexity

Prior to the move day, organizations should freeze their application configurations and not introduce change and complexity.  Any change undertaken during a move adds risk and complicates the project.

But there is one exception to the “no changes” rule: An organization should consider the purchase of new core networking gear before the move. This will significantly reduce the risks and down-time of reinstalling network gear at the new site during the move. The network should be up and fully tested before any move is made.

5 – Plan For Dedicated Resources On Moving Day

Managers undertaking a data center move must create detailed checklists covering everything from server rails to cabinets to specialized tools (e.g., screw threaders for stripped screws; sledge hammers; special power cords to prevent outages when servers cannot be reconnected, etc.).

To limit downtime, a data center move should begin as soon as applications have been shut down, and the move should be scheduled at a time when involved employees are able to fully focus on the move’s success. Trying to “squeeze in” a move on top of full time, daily responsibilities is doomed to failure. It will create a heavy tax on employee morale and unnecessarily prolong the move and potentially impact the business.

“Operational IT knowledge does not always translate into an understanding of how to best move a data center,” stressed Buettner.  “A move requires careful planning, diligent forethought, and impeccable execution.  Having this spearheaded by someone who’s done it before it a must to ensure success.”

Small businesses have found many ways to reduce capital expenditures and operating costs through cloud computing, in which computing is provided by shared resources and software on the Internet on an on-demand model. As such, they continue to move many business functions from on-site servers to the Web.

But “few businesses truly have a cloud “migration” strategy and even fewer have successfully leveraged the cloud in vendor relations,” said Manon Buettner, Principal of Nuvalo, a Bellevue, Wash.-based consulting and brokerage firm specializing in simplifying cloud migration for companies. “We always work to educate in the most simple way possible when it comes to cloud migration, as it can be overwhelming to many.”

Outsourcing To The Cloud
Maybe you’ve ditched the Microsoft Exchange server in favor of Google Apps. That was easy. It’s just like your Gmail account, but for work.

Whether you realize it or not, your business has already begun an important migration.

But few small businesses have thoughtfully managed this transition and even fewer are fully taking advantage of business in the cloud. Using the cloud as a cost reduction tool is increasingly common – reducing paperwork, lowering transaction costs, and investing in less hardware (and with fewer resources to manage it) can yield immediate impact in accelerating your growth. Less understood, however, is how to use the cloud as a business enabler.

Companies like Salesforce.com, for example, continue to push forward. Without installing any software, vendors can manage the entire customer lifecycle or HR processes with greater data detail and accuracy than ever before. No longer do you need a file cabinet of paperwork or indecipherable database for these important business processes. Outsourced to the cloud, your data becomes globally accessible and more secure and redundant – all while saving you real money.

But as the cloud becomes more pervasive, maximizing the value for your business means going beyond Salesforce.com and Gmail. What if the cloud could help you keep your business on the cutting edge? Start thinking about the cloud as more than just a tool you use – anyone you do business with should be in the cloud, as well.

Leveling The Playing Field
Small and mid-sized businesses have long missed out because of their size – they don’t have the budget to buy enterprise-class hardware and don’t have the scale to show up on the radar of innovative software vendors. The cloud is leveling this playing field.

As an executive, your number-one job is to drive revenue by increasing margins and market share – and there are hundreds of hardware and software solutions that could potentially help you achieve that goal. You’re inundated with calls, e-mails, and advertisements from countless vendors. Let the cloud help here, as well.

What if instead of enduring countless marketing pitches, you could get your hands on any IT product you’re interested in and try it within minutes? This is yet another way the cloud can save you time and money.

“The cloud can help you gain access to new solutions and ensure they match your business needs – all without wasting your valuable time,” added Buettner.

But is it really a fail-proof move? Is the cloud really too big not to fail?

To better understand this is to look at how the cloud can fail, and with this, Amazon is a prime example and the growing pains it experienced.

“Failure of the cloud is not simply that a virtual machine can fail,” said Manon Buettner, Principal of Nuvalo, a Bellevue, Wash.-based consulting and brokerage firm specializing in simplifying cloud migration for companies. “The real issue is that the risk is there for losing an entire availability zone, and sometimes multiple zones for reasons such as power outages, connectivity issues, or even natural disasters. But by addressing these for alternatives, risk is minimized and the cloud is then back to its ‘fail-proof’ status.”

Buettner notes that when you get used to the idea that an entire facility can fail, you construct your infrastructure much differently. For a very long time, enterprise IT operated with a no tolerance for failure model. This is well documented from the mainframe days, and even the minicomputer days, she pointed out.

Systems would be designed, and at the basic hardware level they had to be robust enough so that a single system would not fail. We know now this was a fallacy because looking back, we can see that these legacy systems did in fact fail, and when they did, it would be catastrophic for the impacted company.

Thus, we’ve gone from no failure tolerated, to enterprise IT building entire systems that likely will fail. Making the leap from that level of commitment and understanding, to the notion that entire facilities will also, is a larger jump still. Cloud has bridged that gap so now the way IT thinks about failure is a function of how many systems there are, how they operate, what percentage will be impacted by adjacent failure, and the notion that it doesn’t really matter because at any time a whole facility could go down.

Back to Amazon when its US East datacenter went down: Its single availability zone failure took down all of Amazon AZs to some extent. So why didn’t Amazon prioritize making all its systems fail-proof? In some ways they were working to mitigate facility failures that impacted other facilities, so that if one availability zone went down, it wouldn’t create problems in all the other areas that it had dependencies.

So when looking at cloud providers as a measure of how fail-proof – or rather, ‘fail-ready’ they are – determining what your business needs will be is central to understanding what provider is best for you.

One type of provider will model their systems and services after Amazon, and will help you construct your environment to a point that it can really take an entire site outage and still maintain business continuity to other facilities.

Other service providers will do as much bullet-proofing of their facilities as possible to ensure that their cloud stays up and running.

And depending on the type of app being run, its criticality, and who the consuming user-base for the app is, your cloud strategy will have to be in line with the businesses goals.

Thus, there’s no reason why you can’t take both approaches.

But you need to design for contingency. From the technical perspective, it’s about isolating each layer. And the more abstracted from the infrastructure you can get, the better. You could have the greatest cloud approach in the world, but you could also have your DNS so tightly tied in to how things operate that if you lose those servers tied to the DNS, you lose everything.

Hence, the more you can isolate layers of infrastructure from each other, the more you can make them repeatable across multiple facilities.

At times, using technologies outside the scope of the datacenter – a globalized third party DNS built in such a way that it can’t go down, for instance – will help minimize your cost of duplicating such a system. Running real time in multiple facilities, whether the back-up facility is cold or warm, what your data replication model is between multiple facilities, and how much latency you are willing to accept between facility A and facility B are all key determinations in figuring out what cloud approach is best for your application and business.

In the end, just like all other parts of your business, your cloud strategy should reflect the level of risk you are comfortable supporting in your organization.

“The cloud is the way to go, and like any IT infrastructure, it’s certainly not fail-proof,” added Buettner. “But it’s definitely viable when architected correctly.”

DatacenterDynamics Seattle is specifically designed to fill the knowledge and networking needs for both those responsible for the design, build and operation of IT facilities as well as key IT decision makers responsible for strategic decisions regarding capacity planning and technology investment. Key topics will include ‘Design, Build, Operate,’ ‘Outsourcing Decisions’, and ‘IT Optimization.’

For conference details visit http://www.datacenterdynamics.com/conferences/2012/seattle-2012

About Datacenter Dynamics

DatacenterDynamics is a full service B2B information provider at the core of which is a unique series of events tailored specifically to deliver enhanced knowledge and networking opportunities to professionals that design, build and operate data centres.

With 40 established annual conferences in key business cities across the world, DatacenterDynamics is acknowledged as the definitive event where the leading experts in the field share their insights with the top level datacentre operators in each market.