4 Reasons Cloud Storage is Not a Bubble About to Pop

With the recent S-1 filing by Box for their Initial Public Offering (IPO) the question of a Cloud Storage Bubble is raised once again. But is it really a bubble? And should enterprise customers take note and run for the hills? There is more at stake than what appears on the surface.

Box Files Form S-1 IPO

By Box filing their S-1, their financials are put on display for all to scrutinize. Within those figures, we learn that their 34k+ paying customers contribute $124m in revenue that offsets operational costs to the tune of a $169m loss last fiscal year. Over the past four years of reporting, Box reported an increase in the loss trend. But is this enough to consider impending doom?

Cloud Storage Startup Landscape

In 2013, Nirvanix (another cloud storage startup) closed up shop and sent their customers scrambling. Dropbox is another of the closest competitors to Box and announced their intent to IPO as well. Could Box and Dropbox be following in Nirvanix’ footsteps? Enterprise storage is expensive. Yes, there are economies of scale and tricks you can play to maximize the efficiency, but storage infrastructure is expensive.

So, let’s take a look at some potential hypothesis on what may be occurring:

Hypothesis One: There is a minimum amount of capital required to achieve profitability.

Nirvanix only took on $70m while Box and Dropbox took on $414m and $607m respectively. Consider that enterprises need stability in their cloud storage provider, a substantial number of enterprise features (ie: auth, security) and a solid ecosystem for integration. It is probable that $70m is not enough to reach ‘escape velocity’ in this space. It is possible that $400-600m may not be enough either. It is also likely that scale plays a significant role too. It will be interesting to see Dropbox’ figures when they file their S-1.

Hypothesis Two: The real value for cloud storage is not in unstructured file storage.

Sure, the ability to store, share and collaborate on files online is valuable. However, is there greater value in the meta-data that comes from understanding the behaviors of those files? Plus, similar to the problem email systems and enterprise storage vendors addressed years ago with data de-duplication, there is value to managing files at scale. Not to mention that the meta-data around that data could be repurposed for other functions.

Hypothesis Three: Unstructured file storage is simply a loss leader.

There are many directions a company like Box or Dropbox could take based on their current service offerings. Of course there are many directions this could take, but that is for a future discussion.

Hypothesis Four: The shifting enterprise storage paradigm will not allow cloud storage failure.

It is simple enough to treat all storage the same, but in reality it is not that easy. Traditional methods for storing files on internal storage sub-systems is cumbersome at best when we move into a SMAC (social, mobile, analytics, cloud) based world. Enterprises are already shifting toward cloud-based storage to alleviate the pressure and shift their paradigm. The thought of having to move back to traditional methods would break many apps and services. In the end, enterprises really need to move forward and are not able to go back.

Consider the Options

On the surface, it may appear that Box (and ostensibly Dropbox) may be losing money today, there is much more at stake. Enterprises know they need to make a shift to a SMAC based world too. The cards appear to point favorably in the direction of additional options beyond the currently cloud storage portfolio offering. I would look more toward the future opportunities of the space through one of the four hypotheses and less on the impending implosion.

The Shark of HP Converged Systems

The story of Converged Infrastructure (CI) continues to gain steam within the Information Technology (IT) industry…and for good reason. Converged solutions present a relatively easy way to manage complex infrastructure solutions. While some providers focus on CI as an opportunity to bundle solutions into a single SKU, companies such as Nutanix and HP have produced solutions for a couple of years now that go much further with true integration.

As enterprise IT customers shift their focus away from infrastructure and toward platforms, application and data, expect the CI space to heat up. Part of this shift includes platforms geared toward specific applications. This is especially true for those operating applications at scale.

Last week, HP announced their ‘shark’ approach of hardware solutions geared toward specific applications. One of the first targets is the SAP HANA application using HP Converged System 500 as part of a co-innovation project between HP & SAP. It is interesting to see HP partner with SAP HANA with so much emphasis on data analytics today. In addition, specialized solutions are becoming increasingly more important in this space.

Enterprise IT organizations need the ability to start small and grow accordingly. Even service providers may consider a start-small and grow approach. Michael Krigsman (@mkrigsman) recently wrote a post outlining how IT projects are getting smaller and still looking for relief. HP expressed their intent to provide scalable solutions that start small and include forthcoming ‘Project Kraken’ solutions later this year. Only time will tell how seamless this transition becomes.

Additional Reading:

HP CS Blog Entry:

http://h30507.www3.hp.com/t5/Converged-Infrastructure/HP-ConvergedSystem-for-SAP-HANA-meet-the-industry-s-most/ba-p/157176#.UynDsdy0bfM

HP Discover Barcelona: What to Watch For

Today kicks off HP’s Discover conference in Barcelona, Spain with a bevy of information on tap. Looking over the event guide, it is clear that HP is targeting the Enterprise customer with an emphasis on Cloud Computing, Data (including Big Data) and Converged Infrastructure. HP’s definition of ‘converged infrastructure’ does include a bevy of their core infrastructure components.

With an emphasis on cloud and data, HP is really targeting the future direction of technology, not just traditional IT. HP is a large company and can take a bit of work to evolve the thinking from traditional IT to transformational IT. It is good to see the changes.

Of note is the expansion of data beyond just Big Data. For many, the focus continues to persist on Big Data. Yet, for many enterprises, data expands well beyond just Big Data. Look for more information beyond the existing NASCAR example on both the breadth and depth. In addition, there are sessions that provide a deep dive specifically for HAVEn partners. It is good to see HP consider the importance of their partner program.

Core areas of both printing and mobility are making an appearance here at Discover. However, their presence pales in comparison with the big three.

So, what to look for… With cloud and data, the keys for HP will rest with how well they enable adoption. How easy do they make it for customers to easily adopt new technologies? Adoption is key to success. With converged infrastructure, has the story of integration moved beyond a reference architecture and single SKU approach? Look for more details on how far HP has come in developing their portfolio along with execution of the integration between the different solutions. This integration and execution is key.

Time to get on the Colocation Train Before it is Too Late

The data center industry is heading toward an inflection point that has significant impact on enterprises. It seems many aren’t looking far enough ahead, but the timeline appears to be 12-18 months, which is not that far out! The issue is a typical supply chain issue of supply, demand and timelines.

A CHANGE IN THE WINDS

First, let’s start with a bit of background… The advent of Cloud Computing and newer technologies, are driving an increase in the number of enterprises looking to ‘get out of the data center business. I, along with others, have presented many times about ‘Death of the Data Center.’ The data center, which used to serve as a strategic weapon in an enterprise IT org’s arsenal, is still very much critical, but fundamentally becoming a commodity. That’s not to say that the overall data center services are becoming a commodity, but the facility is. Other factors, such as the geographic footprint, network and ecosystem are becoming the real differentiators. And enterprises ‘in the know’ realize they can’t compete at the same level as today’s commercial data center facility providers.

THE TWO FLAVORS OF COLOCATION

Commercial data center providers offer two basic models of data center services: Wholesale and Retail. Digital Realty and DuPont Fabros are examples of major wholesale data center space and Equinix, Switch, IO, Savvis and QTS are examples of major retail colocation providers. It should be noted that some providers provide both wholesale and retail offerings. While there is a huge difference between wholesale and retail colocation space, I will leave the details on why an enterprise might consider one over the other for another post.

DATA CENTER SUPPLY, DEMAND AND TIMELINES

The problem is still the same for both types of data center space: there is a bit of surplus today, but there won’t be enough capacity in the near term. Data center providers are adding capacity around the globe, but they’re caught in a conundrum of how much capacity to build. It typically takes anywhere between 2-4 years to build a new data center and bring it online. And the demand isn’t there to support significant growth yet.

But if you read the tea leaves, the demand is getting ready to pop. Many folks are only now starting to consider their options with cloud and other services. So, why are data center providers not building data centers now in preparation for the pop? There are two reasons: On the supply side, it costs a significant amount of capital to build a data center today and having an idle data center burns significant operational expenses too. On the demand side, enterprises are just starting to evaluate colocation options. Evaluating is different from ready to commit spending on colocation services.

Complicating matters further, even for the most aggressive enterprises, the preparation can take months and the migrations years in the making. Moving a data center is not a trivial exercise and often peppered with significant risk. There are applications, legacy requirements, 3rd party providers, connections, depreciation schedules, architectures, organization, process and governance changes to consider…just to name a few. In addition to the technical challenges, organizations and applications are typically not geared up to handle multi-day outages and moves of this nature. Ponder this: When was the last time your IT team moved a critical business application from one location to another? What about multiple applications? The reality is: it just doesn’t happen often…if at all.

But just because it’s hard, does not mean it should not be done. In this case, it needs to be done. At this point, every organization on the planet should have a plan for colocation and/or cloud. Of course there are exceptions and corner cases, but today they are few and shrinking.

COMPLIANCE AND REGULATORY CONCERNS

Those with compliance and regulatory requirements are moving too…and not just non-production or Disaster Recovery systems. Financial Services organizations are already moving their core banking systems into colocation. While Healthcare organizations are moving their Electronic Health Records (EHR) and Electronic Medical Record (EMR) systems into colocation…and in some cases, the cloud. This is in addition to any core legacy and greenfield applications. The compliance and regulatory requirements are an additional component to consider, not a reason to stop moving.

TIME CHANGES DATA CENTER THINKING

Just five years ago, a discussion of moving to colocation or cloud would have been far more challenging to do. Today, we are starting to see this migration happening. However, it is only happening in very small numbers of IT firms around the globe. We need to significantly increase the number of folks planning and migrating.

DATA CENTER ELASTICITY

On the downside, even if an enterprise started to build their data center strategy and roadmap today, it is unclear if adequate capacity to supply the demand will exist once they’re ready to move. Now, that’s not to say the sky is falling. But it does suggest that enterprises (in mass) need to get on the ball and start planning for death of the data center (their own). At a minimum, it would provider data center providers with greater visibility of the impending demand and timeline. In the best scenario, it provides a healthy ecosystem in the supply/ demand equation without creating a rubber-band effect where supply and demand each fluctuate toward equilibrium.

BUILDING A ROADMAP

The process starts with a vision and understanding of what is truly strategic. Recall that vitally important and strategic can be two different things. Power is vitally important to data centers, but data center providers are not building power plants next to each one.

The next step is building a roadmap that supports the vision. The roadmap includes more than just technological advancements. The biggest initial hurdles will come in the form of organization and process. In addition, a strong visionary and leader will provide the right combination skills to lead the effort and ask the right questions to achieve success.

Part of the roadmap will inevitably include an evaluation of colocation providers. Before you get started down this path, it is important to understand the differences between wholesale and retail colocation providers, what they offer and what your responsibilities are. That last step is often lost as part of the evaluation process.

Truly understand what your requirements are. Space, power and bandwidth are just scratching the surface. Take a holistic view of your environment and portfolio. Understand what and how things will change when moving to colocation. This is as much a clear snapshot of your current situation, as it is where you’re headed over time.

TIME TO GET MOVING

Moving into colocation is a great first-step for many enterprises. It gets them ‘out of the data center’ business while still maintaining their existing portfolio intact. Colocation also provides a great way to move the maturity of an organization (and portfolio) toward cloud.

The evaluation process for colocation services is much different today from just 5 years ago. Today, some of the key differentiators are geographic coverage, network and ecosystem. But a stern warning: The criteria for each enterprise will be different and unique. What applies to one does not necessarily apply to the next. It’s important to clearly understand this and how each provider matches against the requirements.

The process takes time and effort. For this and a number of other reasons, it may take months to years even for the most aggressive movers. As such, it is best to started sooner than later before the train leaves the station.

Further Reading:

Applying Cloud Computing in the Enterprise

Cloud Application Matrix

A Workload is Not a Workload, is Not a Workload

LifeSize Tech Day 2013

Video conferencing trumps audio conferencing! Why you ask? More than 80% of communication is non-verbal. So, why don’t more people use video conferencing over audio? There are a number reasons…read on.

HISTORY

While some may feel video conferencing is passé, I attended LifeSize’s TechDay in Austin, TX and now have a different perspective. Founded in 2006 and later acquired by Logitech, LifeSize is a producer of video conferencing equipment and services. Historically, video conferencing has been relegated to two extremes: 1) Personal 1:1 communications and 2) Fixed and proprietary meeting room systems. And until recently, the only option was the fixed and propriety meeting room systems. Today, 70% of all video conference calls are point-to-point (1:1 or room-to-room). The great thing about personal systems (ie: Skype, Google Hangouts or FaceTime) is the ability to use them across multiple devices in just about any location. While some provide group video conferencing, they are often not as high quality as fixed systems with high-end cameras and high-speed data connections.

INCREASING PRODUCTIVITY

As people look for ways to increase productivity, an increase in video conferencing could provide a useful tool. Picking up on the non-verbal communication helps drive clarity and highlight nuances not otherwise visible with audio conferencing. Plus, we know that team interaction provides a greater opportunity for collaboration and team building. Video conferencing, while not exactly the same as being in the same room as other people, is coming very close. Even mobile solutions are providing an interesting spin on the ability to video conference from just about anywhere. By bridging the gap between the fixed systems and the personal systems, users can start up a video conference as easily as they would with a phone call.

SPECTRUM OF SOLUTIONS

Video conferencing sits within a spectrum of communication solutions and alone is a $3b market with a number of different solutions. The different solutions within the spectrum of communications are:

- Audio Conferencing: Commonly used for group meetings, but lacks the video interaction. Audio is easy to access and only requires a telephone to use. All of the backend infrastructure is hosted.

- Web Conferencing: Web conferencing offers the ability to share screens and present documents in a one-to-many fashion. Some audio collaboration may exist, but only limited video or sharing bi-directional.

- Video Conferencing: Provides the ability to interact with both audio and video. It provides attendees to interact with each other visually. Video conferencing itself spans a wide range of needs from 1:1 personal video conferencing to high quality video required when connecting meeting rooms together.

- Telepresence: Similar to Video Conferencing, telepresence provide a very high quality way for multiple rooms to participate in meetings. Telepresence carries a hefty price tag and is best geared for connecting entire rooms of people together.

LIFESIZE PORTFOLIO

The LifeSize product portfolio covers a wide space from their smaller Passport Series that supports a single high-definition (HD) display to their flagship Icon series which supports Dual HD displays along with a myriad of other features. LifeSize even offers a video Softphone solution too. While many of the solutions require infrastructure on premises to support video calls, LifeSize is starting to offer a Hosted Infrastructure option.

Many of the existing solutions on the market today may use standards to communicate between end points, but they don’t integrate well with competing solutions. That becomes evident if you want to start a video conference session between companies that may have standardized on different solutions. LifeSize has taken a different path by leveraging standards to provide interoperability with other competing solutions.

Two factors govern the success of any given solution:

1)     Interoperability: How well does the solution interact with other devices, solutions and products? Not only is it standards based, but how accessible is the solution to use?

2)     Critical Mass: Unlike the fixed systems of years past, newer systems need a critical mass of users to function well. Think Metcalfe’s Law here: The utility of a network increases at the square of the number of nodes within it. The more users using the system, the more valuable it becomes.

WEBRTC

An alternative and simple option would be to launch a video conferencing session in a browser. Google and others are working on that via the WebRTC movement. Today, the browser of choice for WebRTC is Google Chrome. But hopefully that will span out to include other browsers like Internet Explorer, Firefox and Safari. Will WebRTC replace video conferencing? Probably not as it is not able to “ring” someone.

HOSTED SOLUTIONS

It was a bit disappointing that LifeSize’s efforts are not centered around their hosted offering. At least not yet. We know that the market is moving away from on-premises equipment and my point of view is that LifeSize should move full-steam in that direction too.

Another opportunity might be for service providers to host the solution for small medium business (SMB) clients. It could provide an interesting market to help augment LifeSize’s existing hosted offering. However, at this time, LifeSize explicitly forbids multi-tenant use of their solution.

IN SUMMARY

While video conferencing may have been around for some time, I believe we are just starting a to see its mass adoption. It is in the relatively early stages as behaviors change to accept starting a video call just like one would an audio call. The adoption of personal solutions will help change this behavior and in turn help open up video conferencing more broadly in the workplace.

Today, LifeSize offers a great portfolio of solutions with both good quality at an interesting price point. As their hosted solution develops further, it will be interesting to see LifeSize’s adoption in the marketplace.

The Plumbing of Cloud Computing

Over the past several years, the conversation about cloud computing inevitably comes back to the technology and connections between systems. It includes the systems, storage, network and interconnection components that make up the cloud environment. In essence, the ‘plumbing’ of cloud computing.

If we use the analogy of plumbers and water systems to cloud computing, the pictures becomes a bit clearer. The pipes and water systems that carry the water from the reservoirs that store, carry and deliver the water to homes and businesses are analogous to data centers, systems, storage and networking solutions.

The water itself with its quality, temperature, mineral content and such are analogous to the applications and services that leverage cloud computing as their delivery mechanism.

Do users care about the pipes that carry the water? No. They care about the quality and attributes of the water.

The users who benefit from the applications and services delivered via cloud computing care little about this plumbing. Why? They’re far removed from how the applications and services relate to the individual nuances between solutions.

There are those that believe users should understand more about the underlying technology. That’s like saying that a consumer of water should understand the differences between a 45 degree bend, nipple, pressure regulator and the rest. The consumer doesn’t want or need to know the differences. There are specialists that understand what the consumer wants and knows how to deliver it. They don’t burden the consumer with having to understand the backend.

Even with cloud computing there is a concept of service providers. Sure, the water systems that we plug our homes and businesses into are service providers. In the water industry, it includes those that deliver bottled water to our homes and businesses. Turn on the service when we need it, turn it off when we don’t. Vary the volume, number of instances and locations we want the service to best meet the consumer’s demand. But that doesn’t mean that the consumer has to understand how the water got from the ground (in the case of spring water) and into the bottle sitting in their home/ business.

The bottom line is that we need to separate the different roles focus on the end product. In the case of water, it’s the quality and attributes of the water. In the case of cloud computing, it’s the applications and services that the consumer accesses. Leave the plumbing details behind the scenes to the experts in their field and don’t confuse the roles.

Thanks to Tom Lounibos (@lounibos) and Jake Kaldenbaugh (@jakewk) for the inspiration for this post.

Applying Cloud Computing in the Enterprise

There has been a bit of confusion around the applicability of cloud computing in the enterprise space. Recently, the question has come up as to where/ when/ how/ what cloud applies to enterprises and the challenges that enterprises face when considering cloud. Now, that’s a big ball of yarn to address even before you address the secondary complexities.

Ben Kepes wrote a good article in Forbes responding to comments made by an SVP at HP portraying Amazon Web Services (AWS) as a ‘legacy cloud’ and the reality of the situation. Does it really apply to addressing the enterprise ball of yarn? My point of view: If AWS is a legacy cloud, traditional IT infrastructure must be downright Jurassic. Neither statement is true. Nor does it directly address the reality of the challenges that exist.

In response, Jeff Sussna wrote a good counter missive suggesting that NetFlix is more than just an edge case. Jeff goes on to suggest that current enterprise legacy applications are far from static and IT orgs would prefer not to perform an ‘forklift’ upgrade of their legacy apps into the cloud. I couldn’t agree more…but the devil is in the details as to why.

There are several factors to consider:

  1. Differences in workloads: I wrote a missive 18 months ago about the differences in workloads (A Workload is Not a Workload, is Not a Workload). It’s important to characterize what you have (legacy and otherwise). No two will be the same.
  2. Application of Best Practices: There is a common misconception that how one company leverages cloud will apply directly to others. The thinking being: If NetFlix has success, so will I. I call this the ‘lemming approach’. It may have worked for IT in the past, but will not serve us well moving forward. First, one has to go back and understand point #1 and more importantly understand the reasons the solution was chosen. Which leads to point #3.
  3. Business Drivers: What factors apply when considering different cloud solutions? Aside from the technical merits, there are business factors to consider too. Not everything is about technology. Is there a regulatory or compliance requirement? How would one solution support my business drivers better than the next? While those are just examples, the business drivers are unique to each company.

And when you’re ready to move into the cloud, especially a legacy app, a forklift upgrade is probably not at the top of the list for a number of reasons. Risk, cost, effort just being three of the top ones…but there are many more to consider. What about all of the 3rd party partner connections? What about the interconnections between apps? How will processes and data governance change? As you can see, there are many factors that need to be considered before taking that first step.

For many, the simple thought of moving a legacy app and its tentacles into the cloud can bring shutters. That doesn’t mean it shouldn’t be considered. But it does mean that it needs greater care and consideration than a greenfield application.

In the end, does this mean that enterprises can’t learn from what companies like NetFlix, Zynga, Dropbox and others have done in the cloud? Of course not. It just means that it should not be taken as a cookie-cutter approach and adapted as appropriate. Use the aspects that are relevant for your situation and leave the rest behind. One size of cloud does not fit all. This is especially true for legacy applications.

If this sounds downright hard and potentially not worth the trouble, then the point has been lost. The move needs consideration, planning and quite a bit of preparation. Best to get started down the path now.

HP Aims for the Stars with CloudSystem and Moonshot

Over the past few months, I’ve had a chance to spend time with the HP product teams. In doing so, it’s really opened my eyes to a new HP with a number of solid offerings. Two solutions (CloudSystem and Moonshot) really caught my attention.

HP CloudSystem

HP’s CloudSystem Matrix provides a management solution that manages internal and external resources and across multiple cloud providers. The heart of the CloudSystem platform is in its extendible architecture. In doing so, it provides the glue that many enterprises look to leverage for bridging the gap between internal and external resources. On the surface, HP CloudSystem looks pretty compelling for enterprises considering the move to cloud (internal, external, public or private). For those thinking that CloudSystem only works with OpenStack solutions, think again. CloudSystem’s architecture is designed to work across both OpenStack and non-OpenStack infrastructures.

However, the one question I do have is why CloudSystem doesn’t get the airplay it should. While it may not be the right solution for everyone, it should be in the mix when considering the move to cloud-based solutions (public or private).

HP Moonshot

Probably one of the most interesting solutions recently announced is HP’s Moonshot. On the surface, it may appear to be a replacement for traditional blades or general-purpose servers. Not true. The real opportunity comes from it’s ability to tune infrastructure for a specific IT workload.

Traditionally, IT workloads are mixed. Within an enterprise’s data center run a variety of applications with mixed requirements. In sum, a mixed workload looks like a melting pot of technology. One application may be chatty, while another is processor intensive and yet another is disk intensive.  The downside to the mixed workload is the inability to tune the infrastructure (and platforms) to most efficiently run the workload.

All Workloads Are Not Created Equally

As the world increasingly embraces cloud computing and a services-based approach, we are starting to see workloads coalesce into groupings. Instead of running a variety of workloads on general-purpose servers, we group applications together with service providers. For example, one service provider might offer an Microsoft Exchange email solution. Their entire workload is Exchange and they’re able to tune their infrastructure to most efficiently support Exchange. This also leads to a high level of specialization not possible in the typical enterprise.

That’s where Moonshot comes in. Moonshot provides a platform that is highly scalable and tunable for specific workloads. Don’t think of Moonshot as a high-performance general-purpose server. That’s like taking an Indy car and trying to haul the kids to soccer practice. You can do it, but would you? Moonshot was purpose-built and not geared for the typical enterprise data center or workload. The sweet spot for Moonshot is in the Service Provider market where you typically find both the scale and focus on specific workloads. HP also considered common challenges Service Providers would face with Moonshot at scale. As an example, management software offers the ability to update CPUs and instances in bulk.

Two downsides to Moonshot are side effects of the change in architecture. One is in the creation of bandwidth problems. Moonshot is very power efficient, but requires quite a bit of bandwidth. The other challenge is around traditional software licensing. This problem is not new and seems to rear its ugly head with changes in innovation. We saw this with both Virtualization and Cloud. Potential users of Moonshot need to consider how to best address these issues. Plus, industry standard software licensing will need to evolve to support newer infrastructure methods. HP (along with users) need to lobby software providers to evolve their practices.

OpenStack at the Core

HP is one of the core OpenStack open-source contributors. OpenStack, while a very powerful solution, is a hard sell for the enterprise market. This will only get harder over time. On the other hand, Service Providers present a unique match for the challenges and opportunities that OpenStack presents. HP is leveraging OpenStack as part of the Moonshot offering. Pairing Moonshot with OpenStack is a match made in heaven. The combination, when leveraged by Service Providers provides a strong platform to support their offering compared with alternatives.

When considering the combination of CloudSystem along with Moonshot and OpenStack, HP has raised the stakes from a single provider. The solutions provide a bridge from the current traditional environments to Service Provider solutions.

I am pleased to see a traditional hardware/ software provider acknowledging how the technology industry is evolving and providing solutions that span the varied requirements. I, for one, will be interested to see how successful HP is in continuing their path through the evolution.

How to Leverage the Cloud for Disasters like Hurricane Sandy

Between natural disasters like Hurricanes Sandy and Irene or man-made disasters like the recent data center outages, disasters happen. The question isn’t whether they will happen. The question is: What can be done to avoid the next one? Cloud computing provides a significant advantage to avoid disaster. However, simply leveraging cloud-based services is not enough. First, a tiered approach in leveraging cloud-based services is needed. Second, a new architectural paradigm is needed. Third, organizations need to consider the holistic range of issues they will contend with.

Technology Clouds Help Natural Clouds

If used correctly, cloud computing can significantly limit or completely avoid outages. Cloud offers a physical abstraction layer and allows applications to be located outside of disaster zones where services, staff and recovery efforts do not conflict.

  1. Leverage commercial data centers and Infrastructure as a Service (IaaS). Commercial data centers are designed to be more robust and resilient. Prior to a disaster, IaaS provides the ability to move applications to alternative facilities out of harms way.
  2. Leverage core application and platform services. This may come in the form of PaaS or SaaS. These service providers often architect solutions that are able to withstand single data center outages. That is not true in every case, but by leveraging this in addition to other changes, the risks are mitigated.

In all cases, it is important to ‘trust but verify’ when evaluating providers. Neither tier provides a silver bullet. The key is: Take a multi-faceted approach that architects services with the assumption for failure.

Changes in Application Resiliency

Historically, application resiliency relied heavily on redundant infrastructure. Judging from the responses to Amazon’s recent outages, users still make this assumption. The paradigm needs to change. Applications need to take more responsibility for resiliency. By doing so, applications ensure service availability in times of infrastructure failure.

In a recent blog post, I discussed the relationship cloud computing provides to greenfield and legacy applications. Legacy applications present a challenge to move into cloud-based services. They can (and eventually should) be moved into cloud. However, it will require a bit of work to take advantage of what cloud offers.

Greenfield applications, on the other hand, present a unique opportunity to fully take advantage of cloud-based services…if used correctly. With Hurricane Sandy, we saw greenfield applications still using the old paradigm of relying heavily on redundant infrastructure. And the consequence was significant application outages due to infrastructure failures. Consequently, greenfield applications that rely on the new paradigm (ie: Netflix) experienced no downtime due to Sandy. Netflix not only avoided disaster, but saw a 20% increase in streaming viewers.

Moving Beyond Technology

Leveraging cloud-based services requires more than a technology change. Organizational impact, process changes and governance are just a few of the things to consider. Organizations need to consider the changes to access, skill sets and roles. Is staff in other regions able to assist if local staff is impacted by the disaster? Fundamental changes from change management to application design processes will change too. And at what point are services preemptively moved to avoid disaster? Lastly, how do governance models change if the core players are out of pocket due to disaster? Without considering these changes, the risks increase exponentially.

Start Here

So, where you do you get started? First, determine where you are today. All good maps start with a “You Are Here” label. Consider how to best leverage cloud services and build a plan. Take into account your disaster recovery and business continuity planning. Then put the plan in motion. Test your disaster scenarios to improve your ability to withstand outages. Hopefully by the time the next disaster hits (and it will), you will be in a better place to weather the storm.

The Future Data Center Is… Part II

Last month, I wrote The Future Data Center Is… and alluded to a shift in demand for data centers. Just to be clear, I don’t believe data center demand is decreasing. Quite the contrary, I believe demand is exploding! But how is demand for data centers going to change? What does the mapping of organizations to services look like?

First, why should you care? Today, the average PUE of a data center is 1.8. …and that’s just the average. That’s atrocious! Very Large Enterprises are able to drive that to near 1.1-1.3. The excess is a waste of energy resources. At a time when Corporate Social Responsibility and carbon footprint are becoming more in vogue in the corporate arena, data centers are becoming a large target. So efficiency matters!

Yesterday, I presented a slide depicting the breakdown of types of organizations and (respectively) the shift in demand.

It is important to understand the details behind this. To start, let’s take a look at the boundary situations.

SMB/ Mid-Tier Organziations

Data center demand from SMB and Mid-Tier organizations starts to shift to service providers. Typically, their needs are straightforward and small in scale. In most cases, they use a basic data center (sometimes just a closet) supporting a mixed workload running on common off-the-shelf hardware. Unfortunately, the data centers in use by these organizations are highly inefficient due to their small scale and lack of sophistication. That’s not the fault of the organization. It just further supports the point that others can manage data centers more effectively than they can. Their best solution would be to move to a colocation agreement or IaaS provider and leverage SaaS where possible. That takes the burden off those organizations and allows them to focus on higher value functions.

Very Large Enterprises (VLE)

At the other end of the spectrum, Very Large Enterprises will continue to build custom solutions for their web-scale, highly tuned, very specific applications. This is different from their internal IT demand. See my post A Workload is Not a Workload, is Not a Workload where I outline this in more detail. Due to the scale of their custom applications, they’re able to carry the data center requirements of their internal IT demand at a similar level due to their scale. If they only supported their internal IT demand, their scale would pale in comparison and arguably, so would their efficiency.

Enterprises

In some ways, the VLE without the web-scale custom application is a typical Enterprise with a mixed workload. Enterprises sit in the middle. Depending on the scale of the workloads, characterization, organization and sophistication, enterprises may leverage internal data centers or external ones. It’s very likely they will leverage a combination of both for a number of reasons (compliance, geography, technical, etc). The key is to take an objective view of the demand and alternatives.

The question is, can you manage a data center more effectively and efficiently than the alternatives? Also, is managing a data center strategic to your IT strategic initiatives and aligns with business objectives? If not, then it’s probably time to make the shift.

Related Articles:

Mark Thiele: Measuring the Size of a Data Center – Yes, it Matters

The Green Grid: Metrics and Measurements