It is very easy to see the cloud revolution as a loss of agility and control for the enterprise, especially for the IT department. The argument goes like this: I’ll be locked in to the service provider just like I am to other vendors and third party data center operators – because the cost of change is too high and/or the effort too difficult.
Yet it is very likely that control with agility will actually be enhanced by the cloud, especially for large enterprises already investing heavily in 3rd party data centers. Note: Third parties already house more than 10% of all data center space, per a January 2012 report.
The following are two scenarios where the hybrid cloud (equipped with automation) could significantly increase control.
1) The Legacy Complexity Tax
For many enterprises the soft costs of adding a server into an environment is higher than the actual cost of the server and software. One large company exec, for example, showed me a 30+ step process chart outlining every step required for simply adding more server capacity to the data center. Even worse, an external auditor had determined that even moving an existing server was more expensive than buying a new one. Complexity added a very high tax on agility and growth, even at this leading tech company.
Complexity makes every change more expensive, undermining the business case for innovation and growth. Traditional IT environments are by nature very complex, especially compared to the web-scale IT practices of Google, Facebook and others.
Virtualization succeeded in reducing complexity, especially around hardware, but it often came with a price; agility was often limited to increasingly expensive x86 stacks (requiring vendor lock-in and aggressive price increases). Vendor licensing costs ended up replacing the cumbersome management costs which were building before the rise of VMware. Complexity in traditional data centers combined with the rising costs of virtualization takes us to the second scenario which will drive demand for hybrid cloud adoption.
2) Third Party Data Centers
While data center operators offered more agility and efficiency for enterprises struggling with complexity, many required service contracts that added new costs (and even new processes and procedures) for IT management and maintenance. The third party data center was a step forward for those burdened with rising complexity and entropy, but there were cumbersome strings attached. Many 3rd party disaster recovery providers, for example, charge for tests and/or dry runs and require onerous procedures that ultimately reduce agility.
Some eased the pain by deploying more agile private clouds with virtualization, yet still add external processes and procedures (and costs) A DR dry run that an internal team could once conduct on its own with minimal budget impact can cost $20,000 and higher.
Even with virtualization and the rise of third party specialists, agility came at higher costs and in some cases with procedures that eroded the benefits.
A Cloud Perspective
As public cloud IaaS (led by Amazon AWS) emerged, enterprises were slow to respond. The public cloud had a control and security problem. Even worse, the effort to enter and exit a public cloud threatened just another form of lock-in. Enterprises had been there, done that. IaaS replaces hardware yet risked maintaining the complexity game that sets the stage for price increases and complexity taxes. Yet enterprises want agility which is speed with control.
That is where hybrid cloud automation enters the cloud picture. As hybrid cloud migration processes (and hybrid clouds) are automated the enterprise can control cloud environments as extensions of their own data centers and use each as needed or when optimum. The cloud reduces complexity and delivers potentials for new operating models.
“Users get a range of choice that helps them become more productive and IT gets lower costs and less complexity.”
Soon after the HMG Summit I got to spend some time with Gartner’s Cameron Haight, and we talked about his recent Web-Scale IT preso delivered at the recent Gartner Data Center Conference. He clarified that web-scale was as much about agility and the reduction of complexity than the creation of massive data centers.
As long as a moat of complexity stands between the cloud and the data center, the cloud will not enable the agility of web-scale It requirements, but rather the ability to merely spin up stovepipes faster.
Hybrid cloud integration and automation will enable strategic agility and availability along with increased operating efficiency. The cloud would offer enterprises the option of “leapfrogging” investments in 3rd party racks and stacks by leveraging APIs and services for existing apps with limited complexity. Enterprises could emulate web-scale IT practices without having to make the massive investments today required for re-architecting and developing new apps for the cloud, which increase lock-in risks.
The “pay as you go model” for existing physical and virtual apps is revolutionary, as hybrid clouds would allow enterprises to use and pay for clouds only when they need them. Rather than paying a monthly fixed fee for traditional disaster recovery, with hybrid cloud enterprises could pay a cloud provider only during an outage or when testing. That could reduce disaster recovery costs by more than 50% while giving internal IT teams the ability to initiate tests without onerous 3rd party charges and processes.
Developers similarly could spin up additional workloads quickly and easily and duplicate apps and services for testing and training purposes as needed, without waiting for servers to be added. The cloud becomes an on demand pay as you go environment, enhancing agility and control, versus today’s stereotype of the cloud as just another stovepipe.
This vision of a hybrid cloud isn’t just strategic to the future of enterprise IT, but also to the future of about a dozen new and existing service providers, starting with Amazon, Microsoft and VMware. Hybrid cloud automation is the critical step between traditional IT and private clouds and web-scale IT.