An evolution, not a revolution

Do you remember where you were at 12:01 a.m. on Saturday, August 1, 1981? Don’t tell me you don’t remember the importance of that pivotal date in world history. The change that came that day has revolutionised the whole world – it was the day that MTV officially began and the infamous Buggles ‘Video Killed the Radio Star’ first aired, signalling the death of music on the radio. What a wonderful day that was. I marvel at old timers musing about how they used to listen to music on the radio.

Except, it didn’t really happen that way. MTV wasn’t really revolutionary. It wasn’t even a new idea. And guess what, I listened to music on the radio this morning on my way to work. The fact is, MTV was an evolution of concepts that had been around since the 1960’s when the Beatles’ ‘A Hard Day’s Night’ is viewed as the first use of creative music video on film. This concept was brought to TV multiple times before 1981, on such diverse shows as the Smothers’ Brothers Comedy Hour and the Monkees.

In the 1970s, the creator of the ‘Top-40’ radio format in Philadelphia, Bob Whitney, sold the idea for a music video marathon called THE NOW EXPLOSION to TV stations and it broadcast hundreds of hours of music videos for the 13 weeks that it ran. That show’s directors, Steve Rash and Fred Bauer, formed a new music video company called Innovisions, which created a new show called THE MUSIC CONNECTION. After that show was cancelled, Viacom bought the rights to it and later started MTV… As you can see, MTV wasn’t the inventor of the music television show format, it was simply the earliest long term successful implementation. Nor did MTV ‘kill the radio star’ as the Buggles claimed in the 1981 debut of MTV. What we learned from this was that there was definitely demand for a form of music television service if done right, and that it would complement, not kill, radio. This same lesson applies to the amorphous term, ‘cloud computing’. Early pundits claimed that the revolutionary idea of cloud computing would kill traditional on premise applications and make private data centres obsolete. One look around at the state of both cloud computing and enterprise data centres should give you all the evidence you need to draw parallels between cloud computing and music television. The basic fact is that cloud computing, however you define it, is not revolutionary.

According to the National Institute of Standards and Technology (NIST), ‘cloud computing is a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction.’

This is simply an evolution of earlier capabilities for enterprises to provide similar services to internal customers through the use of mainframe computers. Granted, in the earlier days of mainframe computing, the sharing was confined to the walls of the enterprise given the World Wide Web came much later. However, the concepts for cloud computing were born in these earlier days of computing when price and availability of skilled labour necessitated that computer resources be shared across an organisation and paid for by the divisions based on usage. Okay, maybe there was a revolution in rapid provisioning between the mainframe and today’s definition of cloud, or maybe that was just an uprising of frustrated users who had to wait weeks to even get a green screen terminal placed on their desk.

During that same time, J.C.R. Licklider envisioned the idea of an ‘intergalactic computer network’, which became the basis forARPANET. (If you know your history, ARPANET was the basis for what we commonly call the World Wide Web, or internet, today.)
And then cloud computing ‘burst’ on to the scene when University of Texas information systems professor Ramnath Chellappa first uttered the words in 1997 at the INFORMS conference in Dallas, saying ‘Computing has evolved from a mainframe-based structure to network-based architecture. While many terms have appeared to describe these new forms, the advent of electronic commerce has led to the emergence of ‘cloud computing.’ This work aims at analysing the role of agents and intermediaries enabling this framework.’
The first real shot across the bow of the traditional data centre came in 1999 when Mark Benioff, with backing from Larry Ellison, founded the Customer Relationship Management (CRM) software company Salesforce.com in a small San Francisco apartment and ‘the end of software’ revolution begins. (Ironically, Larry Ellison also financed another ‘cloud’ start-up, NetSuite (formerly NetLedger) a year earlier, but Salesforce.com is widely viewed as the first Software as a Service( SaaS) company.)
The advent of SaaS really did have the potential to disrupt traditional data centres and sourcing patterns. Suddenly, companies had an option to use mission-critical enterprise software that didn’t require any additional computers in the data centre. More importantly, business units discovered Salesforce.com and began directly purchasing CRM ‘seats’ without the knowledge or backing of their IT organisation.

Guess what? IT departments didn’t go away. Instead, they caught on to the ‘rogue applications’ which were appearing throughout the company. Typically this happened when these users realised that in order for their new whiz-bang application to be truly useful, it need to share information with another system which was managed by IT. And so began the integration of SaaS into the traditional IT realm of enterprise computing.
The next major threat to enterprise data centres surfaced in 2002 when Amazon created Amazon Web Services, a suite of cloud-based services including storage and computation, starting the era of infrastructure as a Service (IaaS). This was later made commercially available in 2006 when Amazon released a version known as the Elastic Compute Cloud (EC2) and Simple Storage Service (S3). These services allowed users to rent compute cycles and storage and pay for what they used.

Surely this was the end for the traditional data centre?! The Buggles were right?! But… it didn’t happen that way. Yes, Amazon and the rest of the IaaS industry (Rackspace Cloud, Azure, etc.) took off. This became an easy way for small companies to host applications without building a data centre (or putting a heat generating server in a closet somewhere.) Large companies didn’t ignore this. Some saw it as an ideal way to supplement their existing data centre resources and help smooth the peaks and valleys of computing consumption. This concept became known as ‘cloud bursting’ and allowed enterprises to tap into external computing cycles when a particular application required more than was available on the internal infrastructure.

Throughout these experiences, IT has exposed to the world the shortcomings of these early forms of cloud computing. The primary early complaints were around security of data and continuous availability of resources, issues that enterprise data centres have been dealing with for years and have become quite adept at addressing. As recently as December, 2012, Amazon had a major outage that affected both Netflix streaming customers and Salesforce.com Heroku customers for most of a day. This followed on the heels of an April 2011 Amazon outage which took some customers multiple days to recover from.

More recently, the focus has shifted to private clouds and hybrid clouds. This is an acknowledgement that the traditional data centre is not going away. From the ‘If you can’t beat ‘em, join ‘em’ book, major cloud infrastructure providers such as Citrix, Microsoft and VMware all tout the use of their virtualisation platforms for establishing private clouds in the data centre.

How is this affecting data centre spending patterns? Data centre spending has been on somewhat of a rollercoaster ride in recent years, in part due to the general economic malaise since 2008.
Cost cutting seems to be the phrase of the day in IT, as evidenced by a recent ESG study (ESG Research: 2013 IT Spending Intentions Survey). In addition, analysts expect that a larger portion of the IT budget will go towards cloud computing initiatives, including SaaS and hybrid cloud implementations.

Finally, ESG also reported in the same IT spending intentions survey that an increasing amount of IT budgets will be spent on maintaining existing infrastructure versus investment in new IT projects. Clearly, enterprises are adjusting to the reality of cloud computing and their spending patterns are reflecting that:
1. Reduce cost
2. Invest in more cloud projects (which purportedly contribute to
number 1 above)
3. One could argue numbers 1 and 2 above would result in the
associated budget share shift from new technology investments
to maintenance of existing technology.

At Emulex, we watch these developments closely since they have
an impact on customer buying behaviour. While we don’t believe
that video really killed the radio star, we do recognise that there is room for both video and radio stars and we love (to sell to) both
of them equally.