In both consumer and business technology discussions, we hear a lot about the cloud these days. So what does “cloud” mean within the context of things like Big Data, Business Intelligence and Data Warehousing?
When we talk about data strategies and cloud technology, there are some basic service categories and deployment models that need to be defined and understood. Here’s a quick look at those categories and models—plus an interesting look at the market itself and where it’s going as predicted by the experts.

Cloud Services

  • Software-as-a-Service (SaaS) — Often referred to as on-demand software, SaaS is a software delivery model in which the computing platform (hardware, network, storage and infrastructure), solution stack (software) and pre-built business services (applications) are accessed by using a Web browser over the Internet. A commonly-referenced example you might be familiar with is SalesForce.com.
  • Platform-as-a-Service (PaaS) — Placed between SaaS and IaaS, PaaS is the provisioning of both the computing platform (hardware, network, storage, infrastructure, etc.) and a specific solution stack (software) to provide a platform for developing custom applications. Force.com, for example, allows users to customize and develop applications on top of the SalesForce.com (SaaS) foundation.
  • Infrastructure-as-a-Service (IaaS) — IaaS provides a complete cloud computing infrastructure service (hardware, storage, network, infrastructure, etc.) that allows the customer to install, configure and manage their own software and applications. A great example of IaaS is Amazon Web Services, which allows companies to deploy a virtual data warehouse or data center without the traditional costs of servers and other onsite hardware.

Cloud Deployment Models

  • Public Cloud
    In the public cloud model, the provider makes the computing platform available to the general public over the Internet. Typically, this deployment model is a pay-per-usage (metered) price structure.
  • Private Cloud
    The private cloud is cloud-based hardware operated solely for a single organization and is often managed internally by that organization. It thus has a few major drawbacks. The private cloud model doesn’t take advantage of the less hands-on nature of cloud services, since only internal personnel administer it. It can also be considerably more expensive, which reduces the typical big cost-saving benefits of a cloud approach.
  • Virtual Private Cloud
    The virtual private cloud exists within the public cloud domain, and while being “virtually” separate from the other cloud customers, the hardware still lives within the public pool of cloud resources.
  • Hybrid Cloud
    The hybrid cloud model connects two or more clouds that can be a blend of public and private cloud models. This model actually includes a few variants. One example connects a traditional on-premise network or server to a cloud-based environment. This is what we see most often when companies begin to utilize cloud options but require connectivity back to legacy systems that cannot be moved to the cloud.

Now that you have a basic idea of the different types of cloud services and deployment models, let’s look at the reasons for migrating projects like Big Data and Business Intelligence into the cloud.

The Benefits of the Cloud

Does a cloud-based data strategy have advantages? You bet it does. There’s a reason so many CIO and IT executives are budgeting for cloud Big Data and Business Intelligence solutions in the near future. Most of the available cloud solutions on the market will render lower costs, scalability, reliability, security, time to focus on your core business, and computing power.

  • Lower Costs
    Cloud computing brings lower costs to Big Data and Business Intelligence right from the start-up and continuing through implementation and ongoing use.

    • Lower Upfront Costs
      Let’s face it, data centers cost a lot of money. With cloud-based Big Data and Business Intelligence, all of the hardware and infrastructure are provided by a service provider. This means the provider bears the majority of the expense that you would typically incur procuring and setting up your own infrastructure.
    • Lower Ongoing Costs
      There’s a lot that goes into the care and feeding of a data center or on-premise solution that often gets overlooked when crunching the numbers. The savings add up very quickly with cloud-based Business Intelligence. Those costs start with the physical space and utility cost. Then there’s the maintenance, support, security, backup and disaster recovery. All of that is typically covered by the service provider. And what about those costly “floor-sweeps” — the periodic replacements of hardware that vendors often recommend or require? Or simply keeping your hardware up-to-date and current as hardware speed and power continue to increase (remember Moore’s Law?).
  • Scalability
    Growing your data as big — or as small — as your company needs it to be is liberating. If you have limited physical space and servers with limited storage capacity, you might be missing out on an opportunity to store precious data. Cloud-based solutions allow you to scale up or down with greater flexibility, because all the computing resources are centralized. You can tailor your data needs to fit your business size today and tomorrow.

    • Unlimited Scalability
      Successful Business Intelligence and Data Warehouse solutions tend to be victims of their own success. Once users begin to understand the value and advantages to be gained through their system and data, usage rapidly increases. Traditional solutions to this challenge have included: proactively over-scoping hardware in anticipation of rapid solution adoption, which, depending on true system success, is a risky spend; or reactively responding with more hardware capacity — a slower option that risks user frustration.
      The cloud solves this challenge by letting you “pay for use” and scale as needed with little to no pain. With database storage, CPU cores and servers, everything can be elastically extended as usage and needs dictate.
    • On-Demand Scalability
      Imagine you have a typical data warehouse load time-frame of 10 hours but your process requires almost all — if not all — of that time. Even if you are lucky enough to have some buffer today, as your system grows in depth and breadth, what may have been ample load time today can quickly shrink to a bottleneck tomorrow.
      With a pay-for-usage cloud model, it’s easy to temporarily scale up your data-load horsepower. For example, at the start of your 8 p.m. load time, a single persistent ETL server can start an extended server farm of unlimited capacity (five, 10, 50 or more servers). You leverage the elastic nature of cloud computing environments to reduce your data-load duration to a fraction of the traditional “on-premise” model. This amazing solution can be even more affordable because you don’t pay for servers sitting idle in a data center during non-load work hours.
  • Reliability
    Most cloud providers have state-of-the-art redundancy and high-availability Service Level Agreements (SLAs) that, honestly, are far beyond the capability of typical companies. The majority of cloud providers also have geographic load-balancing, auto-detecting fault tolerance and real-time backup and recovery. These features are all built right into leading cloud-based solutions. If you want to test a vendor’s merit, those are things to check.
  • Fail-Over/Disaster Recovery
    Even when your on-premise Big Data, Business Intelligence or Data Warehouse solutions aren’t going anywhere, the ability to replicate your environment entirely and provide an extremely reliable fail-over or disaster-recovery option is very valuable. When you combine this capability with pay-for-use and starting the hardware only when it’s needed, you combine the lowest cost with the most reliable technology, which results in an extremely attractive solution option.
  • Security
    Believe it or not, most cloud providers have security requirements that exceed what is typically feasible for midsize, and even very large, companies. How many points of entry are on a typical corporate network? How many employees, contractors and visitors have access to that network? Is there wireless access that can be compromised? Serious cloud providers have some of the most battle-hardened physical and technical infrastructures in the world. No one can simply walk up and “plug in” to your cloud. That can’t be said for many corporate environments.
  • Focus on Your Core Business
    One of the biggest benefits companies cite after implementing solutions in the cloud is the additional time they’re able to spend on their core business. Estimates range up to 80 percent for the time and cost that IT departments spend supporting systems and applications instead of supporting business by developing and deploying useful technology solutions. Using a cloud infrastructure frees up analysts’ and developers’ precious time, and businesses get a better return on their IT dollars.
  • Freedom to Test, Learn and (yes…) Fail Faster
    Big Data and Business Intelligence is often a “what-if”-based endeavor. It often translates into the need to perform one-off or proof-of-concept projects on data to test a business theory or to support a short-term business need. Traditional on-premise business data solutions struggle to allow and support these fast-paced business requirements. Cloud solutions — with rapid provisioning and simple pay-for-use cost models — support these projects. Turning to the cloud opens up entirely new opportunities for using your business data.

The Power of Cloud Computing

Cloud computing means greater power for Big Data and Business Intelligence in at least two crucial ways: It makes powerful data manipulation affordable and accessible for companies of nearly any size.

  • Big Data for All
    While Big Data is nothing new, the cloud makes the ability and ease with which you can leverage that data for business advantage a game changer. With technologies like Hadoop and MapReduce, companies of all sizes can now afford to manipulate and analyze terabytes to petabytes of data, using the virtually unlimited distributed power of cloud computing.
  • Super Computers = Super Data Mining
    Recently, Amazon demonstrated its ability to create the world’s 42nd-fastest supercomputer by launching a virtual supercomputer in their elastic cloud. The specs were impressive: 240 trillion calculations per second, or 240 teraflops on 17,000 cores. But the truly amazing part was the cost — $1,279 per hour. Any business with the need to execute complex analyses or advanced data mining Big Data suddenly has unprecedented processing power and options.

The Challenges of the Cloud

As with every technological innovation, there are challenges. And cloud-based Big Data and Business Intelligence solution isn’t without its woes. But knowing what to watch out for and what to avoid makes the transition to cloud-based projects an easier and less-risky process. Here is a quick look at some of the common challenges that companies big and small face when ascending into the cloud.

  • Data Transfers
    Transferring existing data sets into the cloud can be challenging, especially when they’re big. After years or even decades of doing business, your company may have terabytes of stored data. Data transfer that large is a feat that would choke any system. But fear not. There are new data transfer technologies, like Fast and Secure Protocol (FASP), which allow 2-100x increases in bandwidth utilization. Additionally, most cloud vendors will support physical media transfers, allowing easy one-time migration of your existing solution (terabytes of data) up to the cloud.
  • Buy-in and Support from IT
    Folks in an IT department can be resistant to changing or exploring business technology. Even though companies sometimes added new Big Data or Business Intelligence solutions, without buy-in from IT, these systems were underutilized. But when it comes to cloud technology, opinions are changing and resistance is disappearing. Most IT executives understand that the cloud is the future, and they are planning accordingly.
  • Picking the Right Cloud Provider
    Several vendors are offering some form of cloud solution, and they claim their solution is great. Some of it will be true, but many options won’t be optimal for your needs. Moving to an entirely new infrastructure or platform is never easy. That’s why it’s important to do your homework. Making the wrong choice with a service provider can multiply your data and business challenges while removing some of the advantages outlined in this paper. Review your options and choose carefully.

What the Experts are Saying

If you don’t have a convincing cloud story yet, it’s time to wake up and smell the coffee. That’s the warning from IT leadership group Gartner, which in 2013 released new research that shows rapidly growing acceptance of the public cloud in the enterprise. Gartner predicts continued strong growth in public cloud services, with end-user spending on public cloud services likely to grow 18% in 2013 to hit a global figure of $131 billion. By 2015, that same market could be more than $180 billion.Source:http://www.informationweek.com/cloud/infrastructure-as-a-service/gartner-tells-outsourcers-embrace-cloud-or-die/d/d-id/1110991

“Piper Jaffray enterprise software analyst Mark Murphy today offers up the results of a survey of 141 chief information officers regarding their use of cloud computing, though the details of the time frame in which the survey was taken are not provided. Murphy arrives at the conclusion that only 9.7% of all computing workloads are running in so-called public clouds, of the sort Amazon.com provides with its Amazon Web Services. But that is expected to rise to 30.2% in five years’ time. That adds up to 44% annual growth in workloads for the public cloud versus 8.9% growth for on-premise computing workloads. Source:http://blogs.barrons.com/techtraderdaily/2013/10/17/salesforce-google-amazon-cloud-winners-says-piper-microsoft-straddles-the-line/

According to networking giant Cisco, it’s a cloud, cloud, cloud, cloud world as cloud-based network traffic explodes. Besides data center/cloud network growing at a remarkable rate, Cisco also foresees the majority of server workloads moving from traditional servers to the cloud in 2014. In the third annual Cisco Global Cloud Index (2012–2017)issued on October 15, 2013, Cisco forecasts ‘that global cloud traffic … is expected to grow 4.5-fold.’ That amounts to a 35 percent combined annual growth rate (CAGR) — from 1.2 zettabytes of annual traffic in 2012 to 5.3 zettabytes by 2017. Overall, Cisco expects ‘global data center traffic will grow threefold and reach a total of 7.7 zettabytes annually by 2017. Source: http://www.zdnet.com/cisco-projects-data-center-cloud-traffic-to-triple-by-2017-7000021985/

As public cloud computing gains greater adoption across enterprises, there’s an increased level of spending occurring on infrastructure-related services including Infrastructure-as-a-Service (IaaS). Enterprises are prioritizing how to get cloud platforms integrated with legacy systems to make use of the years of data they have accumulated. From legacy Enterprise Resource Planning (ERP) to Customer Relationship Management (CRM) systems, integrating legacy systems of record to cloud-based platforms will accelerate through 2016.Source:http://www.forbes.com/sites/louiscolumbus/2013/02/19/gartner-predicts-infrastructure-services-will-accelerate-cloud-computing-growth/

Conclusion

There you have it. The basics, the advantages, the challenges and some trends from the experts. The cloud is here and it’s ready for business now. If you haven’t been asked how your Big Data and Business Intelligence systems can take advantage of the options and advantages afforded by cloud technology, get ready because the wave is coming.
By understanding the advantages and challenges, knowing what your options are and moving quickly to test the waters, you can ensure that you are positioned to benefit from this new and potentially disruptive trend.

About iOLAP

iOLAP (www.iolap.com) is a leading professional services and consulting firm that focuses on Big Data and Business Intelligence. We understand you have already invested significant money and time implementing your existing strategic data systems. We help you make those systems better, tune them to their highest potential and then take them to a new level. Industry-changing new projects are on the horizon and we are here to help. New technology shifts are exciting, but they often add confusion to your company’s data strategy. If you need expertise in data system architectures, technologies and methodologies for the development of these new systems, we can help.