Blending Clouds: Avoiding Lock-in and Envisioning the Cloud Fabric
On Wednesday, Feb. 25, rPath will host a webinar with VMware and Bluelock on a topic that has become hotly debated in the chronicles of the cloud: The risk of vendor lock-in. The concern? Enterprises who bet the ranch on a single cloud or virtualization platform could find themselves in an uncomfortably dependent relationship with a single vendor. As I see it, the specific concerns include:
Pricing leverage – where organizations lack a competitive counterbalance to maintain a rational equilibrium in what they’re expected to pay. You’ve seen this before. Ever notice the price of Milk Duds at the movie theater? That’s the familiar form of the same local monopoly caused by real or practical barriers to choice.
Switching costs – where applications and data have become calcified, immovable artifacts of a single platform, making it difficult or impossible to switch. Since virtualization and cloud tends to reduce friction of application deployment, this problem can increase exponentially as a function of time.
Inability to exploit comparative advantage – where organizations are unable to move workloads across clouds to take advantage of comparative advantage in price (think: cloud arbitrage), geographic proximity (think: network latency), and specific capabilities (think: dev/test vs. commodity compute vs. enterprise SLA).
The reality is that a single cloud platform has the potential to become an albatross.
Thorsten von Eickenat at RightScale make an interesting and logically sound assessment of cloud lock in. Thorsten argues that the risk of lock-in increases dramatically in higher layers in the stack: Software-as-a-Service (SaaS) is complete lock in; Platform-as-a-Service (PaaS), slightly less so; and, Infrastruture-as-a-Service (IaaS), more flexible still.
This delineation certainly makes sense, but I believe there is genuine—and warranted—angst about infrastructure lock-in. Ultimately, organizations want to seamlessly blend any combination of datacenters, internal and external clouds into an integrated fabric that allows application workloads to run wherever it makes most sense. The result is an integrated series of clouds—or hierarchies of clouds—that become more like a utility grid than a specific compute target. Policies dictate how and where applications can run and how data and security is handled. The result is a federated environment that liberates organizations from lock-in and provides unlimited scalability, flexibility and range of choice.
At rPath, we automate the creation, deployment and management of application images that can run in any traditional, virtualized or cloud-based environment. Applications packaged with rPath are self-contained and set free to run literally anywhere. You can imagine the implications of this model for the hybrid cloud—a single application capable of running on virtually any server, hypervisor, or cloud.
That’s exactly what we’ll discuss—and demonstrate—during the webinar on Feb. 25th. You’ll see a live demonstration of a real application moving seamlessly between Amazon EC2, Bluelock and a VMware ESX host. And we’ll discuss the tradeoffs and alternatives between cloud options as a basis for envisioning your cloud fabric. We’ll also discuss the role of Open Virtualization Format (OVF) as a standard for application portability, and hear from VMware on their vCloud vision.
It promises to be a great event—and an important one for those with an interest in understanding the risk of lock-in and the promise of the multi-cloud future.
Comments
Leave a Reply
