Quantcast
Channel: iTWire - Entertainment
Viewing all articles
Browse latest Browse all 4710

A CxO guide to the cloud

$
0
0
A CxO guide to the cloud

“The cloud” is the buzzword of the modern day. Companies are told they must “get on the cloud”, while the deeply technical bemoan how what is essentially a hosted service became a special term. If you’re a CxO, here’s the nuts and bolts of the cloud.

“The cloud” as a term has gained significant prominence in recent years. It’s a popular buzzword but despite being spoken of and written about daily, its real meaning and value may not be understood. The cloud is more than a contemporary fad and it is a foundational item that must influence any company’s forward IT strategy.

This is particularly timely, as we enter the second half of the financial year and enterprises begin reviewing their roadmap and multi-year plan. If the cloud is not influencing your direction then it’s time to pause and consider.

{loadposition david08}

Origin and meaning

Historically, companies would host their own data centre. This is the location where the company’s servers, storage, backup, and other core infrastructure facilities were stored. What is considered a “data centre” might range from small and trite such as a server under the receptionist’s desk without power protection, cooling, and various other hygiene and security protections, to large installations with redundant power options, physical security, and raised floors to run cabling through.

Realistically, a small to medium enterprise cannot establish a high-quality, 24/7 physically protected, air-conditioned, dust-proof, redundant-power-supplied data centre due to the cost involved. Two options became available: co-location and hosted. In a co-location scenario, the company pays for rack space in someone else’s data centre and installs its servers there. In a hosted scenario, the company pays for access to someone else’s servers within a data centre, and manages the applications and possibly the operating system on it, but does not own any of the hardware.

As a company moves from self-hosting through co-location through to hosted it can focus less and less on ownership and maintenance of physical hardware, and more on leveraging provided technology to solve business problems.

Within a network diagram, the private network enclosing the company over each of its locations was depicted using a cloud shape. This represented all the locations were inside the one network, which was separate from the outside world.

This is the origin of the term “cloud” and at its most basic, “the cloud” simply means “someone else’s server.”

The cloud today

Technically, it’s true - the cloud is someone else’s server. Yet, it means much more than this, and this understanding will drive business improvement, efficiency gains and cost reductions.  Advances in technology mean dedicated public cloud providers like Amazon Web Services (AWS), Microsoft Azure, Google, Oracle, IBM and others can provide services, with these positive characteristics:

  1. Scale up and down on demand as your workloads change. You pay for the compute power and storage you require, expanding and contracting as needed, without lock-in to any specific scale.
  2. Multiple geographic points of presence, replicating your infrastructure and applications around the globe, giving fast access to users within those regions, and providing high availability in the event any single geographic region experiences an outage.
  3. “Fail fast” and “fail cheaply” options, where you can experiment by rapidly spinning up a server, testing an application or product, and destroying it when done. There is no costly server purchase and installation, no dedicated internal resource to manage it, and experiments can be performed rapidly and inexpensively.
  4. Serverless applications, where databases or websites or other applications can be hosted without any requirement to deal with the underlying server or operating system because it is all managed by the cloud host. Your database or other application still enjoys the benefit of elastic scalability and geographical redundancy but becomes simpler to deploy and manage by the company.
  5. Programmatic construction and destruction of servers, where setting up new infrastructure need not be performed by human hands but can be scripted by a software developer. An environment can be constructed for testing, then perfectly reconstructed for production, or an environment could be replicated thousands of times, every single time flawlessly configured through the accuracy of scripting, then rapidly disposed of when no longer required.

Additionally, the cloud has come to mean web-based applications which are “responsive”, meaning they scale to suit the display device, be it a desktop computer, laptop, tablet or smartphone. The one application works on all devices, and provides access to your data from any device, from any location, at any time, provided you simply have an Internet connection. This is a far change from operating specific fat client desktop applications which could only be executed from within the company office.

Challenges

It’s not all good news. The cloud is not without challenges. The two most immediate of these are:

  1. Security, because your applications are no longer contained within your company firewall but are Internet-accessible. To counter this, public cloud providers invest significant resources into security and maintain just as they can provide a vast installation of redundant power and cooling beyond an affordable price to a “regular” business, so too they provide a vast amount of security beyond what may ordinarily be affordable to a business. Today Fortune 500, Government and Military organisations embrace the cloud, and have put their trust in it.
  2. Cost, because subscription-based cloud services come with a myriad of options. The number of hours your servers are running, the processing power, storage space, storage speed, number of redundant regions, ingress and egress bandwidth, elastic scale, and other factors all affect the price. It is important to continually review cloud infrastructure provisioning against requirements to optimise cost.

Options

Cloud architecture is highly desirable. However, not all companies are comfortable with hosting in the public cloud. You can still gain benefits of cloud scale and elasticity in different configurations, albeit at differing levels of cost and resourcing.

  1. Public – full use of a public cloud, such as AWS or Azure.
  2. Private – maintaining a data centre within a company, but deploying a cloud-based environment over it such as OpenStack instead of the traditional Microsoft Windows Server setup. OpenStack is a mature open source private cloud environment that offers the advantages listed above, albeit the company must still purchase and manage its own infrastructure and the necessary resources to manage it.
  3. Hybrid – whereby a company will deploy some workloads to the public cloud, and some to a private cloud, depending on factors like sensitivity.
  4. Cloud On-premises – this is an option provided by Oracle, who will install and maintain a data centre within the company’s environment, running their very same public cloud environment, applications on it and security, however, it is all contained within the private on-premises environment. This environment still offers scalability with Oracle providing more servers and storage as required. The price is extensive but is an option for large enterprises and Government who have this preference.

How to get there

Unless your business is a startup, it’s unlikely your path to the cloud is simple and without effort. Here are four strategies you can apply to your current situation to assist:

  1. Rehost, lift and shift. This means to take a server from one place and move it to another. This is the very core of cloud server management and it is a solved problem to move workloads from point A (i.e. your data centre) to point B (i.e. cloud provider) using third-party tools that have competed themselves down to a reasonable commodity price point. According to research this has a proven 20% overall savings when you do a complete TCO comparison to running these applications in the data centre, taking into consideration a five-year, forward-looking model that accounts for capex of hardware growth and refresh along with operational overhead of real estate or hosting fees, power/cooling, headcount and software/hardware support contracts.
  2. Replatform. This means to rebuild the server or application, reloading it onto a new server or operating system (for example, Linux), or convert a database to a new engine. Again, research indicates this offers a greater than 20% saving as you shed more operational overhead and/or licensing costs.
  3. Repurchase. Simply put, one strategy is to buy a new application with more features and flexibility. The cost here is less about migration and more about the cost of changing the business to leverage the new application to service the same business need. As an example, this is the strategy you might employ to move from an internally hosted desktop ERP system to a modern cloud-based product like NetSuite or Microsoft Dynamics NAV.
  4. Refactor. If your applications are in-house you have the option of rewriting a legacy application to use cloud-based services.

The last business architecture / the new normal / innovation

For your enterprise’s IT roadmap, the cloud is highly relevant.

  1. Accessibility – you need your core applications to be available anywhere, anytime, on any device. This includes senior management, regional and site managers, and potentially your customers themselves.
  2. Reliability and resilience – If your organisation currently owns and maintains large numbers of site-based servers I would be surprised if without exception there was not at least one server inaccessible for some period every month, be it local power failure, server hardware failure, local network failure, or some other reason.
  3. Re-allocation of spending – also without exception, every month, there are undoubtedly costs incurred to replace elements of the legacy environment of these on-site servers, be it a new backup drive, backup media, hardware warranty renewal, after-hours charges for outsourced providers to investigate faults and reboot servers, or some other item. Anecdotally, research indicates 80% of an enterprise’s ICT spend goes on maintaining legacy environment, with only 20% on innovation. The cloud provides an opportunity to spin this around, such that 80% of spend can be on innovation, by removing the burdens of owning and maintaining hardware. This is the value proposition your IT roadmap can deliver.

I have been privileged to attend world-class conferences listening to the insights of very large organisations and observing worldwide trends. Some notable quotes:

  1. Zach Nelson, at the time CEO of NetSuite, stated “the cloud is not only the next business architecture, it is the last business architecture. Once you can access your data on any device, at any time, from anywhere, what else is there?”
  2. Andy Jassy, CEO of AWS, stated “the cloud is the new normal.”
  3. Mark Hurd, CEO of Oracle, stated “the cloud is not if, but when … you get innovation straight away … it’s more secure."
  4. Hurd further predicts by 2025 80% of production applications will be in the cloud, and that enterprise clouds will be the most secure place for IT processing. “Companies are moving quickly. Don’t get left behind,” he stated.

 


Viewing all articles
Browse latest Browse all 4710

Trending Articles