If you read the avalanche of marketing material from venture-funded loss-making startups in the cloud space you’d think that on-premise infrastructure will become akin to old typewriters lying around in store rooms and that IT departments will go the way of the dinosaurs after the asteroid hit. Any vendor advocating anything else is toast. Geddit?
We all know that controversy brings eyeballs and that a wild claim can be somewhat off the mark but still indicative of a major disruptive change in the market. In 2007 Mark Benioff said that “Software was dead!” which he later admitted was an exaggeration on his real point that the model for software was ripe for change. The reality was that we were going through a transition in the market where the opportunity was around new models and services; a transition that witnessed the way that we consume, manage, build, and deploy software radically change from the days of packaged licensed software, multi-year upgrades and slow innovation cycles. He was right and software is thriving.
We’re going through a similar transition in cloud and, as we shift into the next wave, the driver for cloud adoption is shifting from economics-led to innovation-led as leading-edge companies invest in cloud services as the foundation for digitizing their companies and, in so doing, creating new forms of customer value. While the model in enterprise IT is changing, this will in turn lead to a second bigger wave of cloud adoption that spans both the data center and the public cloud.
When we launched Intercloud in March of 2014 we said it would be architected to enable hybrid cloud, security, data sovereignty, management control, and services choice. It would also leverage the huge assets of our partner ecosystem in a way that enabled them to profitably participate in the model. All of the external market developments we have witnessed in the last two years have re-affirmed that these design criteria were not only correct but have now become the key criteria for the next wave of cloud enabled industry disruption – Digitization.
To see this consider the four areas in particular that will determine success in this next phase:
Cloud Battleground 1: Security from the Data Center to the Edge
As the Internet of Things (IoT) grows we will see a convergence of customers’ existing information technology (IT) and operational technology (OT) networks. This convergence significantly expands the security challenges, due to its increased breadth and depth over existing network connectivity. The IT and OT networks are managed with different priorities in mind, and each has distinct security needs. Today, the priority of the IT network is to protect data confidentiality and the focus of the OT network is on physical security and secure access to ensure operational and employee safety. With the convergence of these two environments and the hyper distribution of applications across hybrid cloud architectures, IoT security requires a new approach that combines physical and cyber security components driven by cloud-based software analytics that detects anomalies and takes pre-emptive action.
Cloud Battleground 2: Orchestration and Multi-Cloud Management
The explosion of cloud services will drive a significant need for scalable, automated and policy-driven orchestration with multi-cloud management from the data center to the edge, for networks, operations, applications and all IT policies. This orchestration needs to enable low-touch on-demand provisioning for both network and operations with compliance that spans regulatory and security requirements for stringent service-level agreements (SLAs). As a subset of orchestration, multi-cloud management will become one of the most strategic control points for cloud as customers will require single pane of glass management capabilities that provide visibility, automation, control and freedom of choice over all cloud services accessed across the Enterprise. The need for this is already apparent as our Cloud Consumption as a Service data shows where the average company is now accessing over 1,200 cloud services but the IT department has approved less than 100.
Cloud Battleground 3: Developers and the Next Generation PaaS
Recently we have seen managed OpenStack solutions, such as Cisco Metapod and associated platform as a service (PaaS) tools, enable a new more agile DevOps approach to software development of cloud native applications. Christened the ‘Modern IT Stack’, this is enabling the new DevOps movement in enterprises where new application development and deployment are integrated into the same process without the traditional handoff between developers and deployers. Containers have become the preferred deployment model for cloud native applications. By containerizing the application platform and its dependencies, differences in OS distributions and underlying infrastructure are abstracted away and the application becomes more portable. This makes them ideal for the hyper distributed IT models we see emerging where applications take the form of microservices accessing data at the edge. This means that the future hybrid cloud platforms need to embrace all three models – bare metal, OpenStack and containers – and successful vendors must win the software developers mindset to develop on our their cloud platform regardless of which tools they choose. After all, Stephen O’Grady said it well “developers are the new kingmakers.”
Cloud Battleground 4: IoT and Analytics: Data Is the New Oil
As 50 billion things connect to the network and harnessing the relevance of this data becomes vital to business process digitization, companies must begin treating data as an enterprise-wide corporate asset while also managing the data locally within business units. The challenge is that data at the edge is not only many times bigger making ‘extract, transform and load’ techniques slow and expensive but the data itself is ephemeral. It needs to be captured, only the relevant portions moved, and then given context by analyzing it with other relevant data elements—where they are, what they are, what state they are in, and real time conclusions drawn on what all this data means for business decision making. This will give rise to ‘data developers’ who will use machine-to-machine (M2M) capabilities to define and access data analytics services on an ‘IoT Fabric.’ Every company will require a cloud-based IoT Fabric that will be a conduit through which IoT data services will be securely defined, accessed and managed. The end result and its capability to refine data into greater business insights, more profitable decision-making, and new product offerings is clearly one way to make a substantial impact on the business.
These developments show that the second phase of cloud will be bigger than the first. At Cisco, our estimate is the TAM for solutions that meet these four criteria is around $200 billion over the next few years of which only around $30 billion will be addressed by what we call public cloud today.
That’s a lot of typewriters…enjoy your toast.
of physical space. You might say ‘we are building a new data center that is sure to offer higher density’ or ‘we are moving to a new co-location facility and this won’t be an issue.’ However, it takes expensive and complex technology to achieve higher average rack cooling density on the facility side, and critical trade-offs on the server side….
