Yet another new addition to your buzzword bingo card: "Pluggable cloud" is defined by Deloitte Consulting's David Linthicum as the strategy of moving workloads from one public cloud to another to get the best price, or otherwise get the best deal.
"Pluggable cloud" treats Amazon Web Services Inc. , Microsoft Azure , Google (Nasdaq: GOOG) and their infrastructure and platform competitors like interchangeable components. Like light bulbs. Or like rooms in business-class hotels. Stay in Hilton on one trip, Marriott the next trip, Hyatt the trip after that -- whatever hotel has the best rate when you're traveling.
They're all the same.
The enterprise "pluggable cloud" strategy takes a similar view of pluggable cloud providers, says Linthicum.
"A pluggable cloud is a multicloud setup where you can swap out the public or private clouds without having to change much of the underlying application dependencies," he says. Enterprises change out clouds "to deal with price and functionality changes," he says.
Linthicum is skeptical that this is even a thing. It requires a cloud service broker (CSB), cloud management platform (CMP) "or other tools that provide abstraction away from the cloud native services," Linthicum says. The alternative is managing each cloud service independently.
And it takes a lot to get everything to work right. For example, storage implementations are different among different public cloud providers, Linthicum says. "Even if you do a great job creating abstraction and orchestration layers, there is a great deal of work needed to make it actually work. I'm not sure 'pluggable' would be the word I'd use," Linthicum says.
Even if you can get it to work, you're forced into a least-common-denominator approach, using only the basic functions for the cloud providers, rather than the full range of services that are unique to each provider, Linthicum says.
Linthicum makes good points here, but he also suggests a false dichotomy. You don't have to choose between pluggable cloud and committing heart and soul to a single cloud provider. Pluggable cloud can be a component of an overall enterprise cloud strategy. Take full advantage of each cloud provider's specialty -- Amazon's Lambda serverless computing, or Google's artificial intelligence and machine learning -- where it makes sense for your business. And, for other workloads, use services like Kubernetes and CLoud Foundry to keep your pluggable cloud freedom.
Alternately, you can give nearly all your business to a single cloud provider, while keeping some workloads on alternative providers, just so you're aware of what's out there, and so your primary cloud provider knows you have the freedom to switch.
Pluggable cloud is an aspect of a multicloud strategy, where enterprises split their business among multiple cloud providers, for platform and infrastructure -- where Amazon, Microsoft and Google are the US market leaders -- as well as for applications such as Office 365, Salesforce, Oracle and SAP.
Cloud providers and infrastructure vendors are lining up on both sides of the multicloud issue. The pro-multicloud faction includes Microsoft, Google, Cisco and VMware; pretty much everybody but Amazon and Oracle at least give lip service to the multicloud idea. (See Google & Microsoft Tout Multi-Cloud, but Where's Amazon? and Following Amazon Partnership, VMware Is a Cloud Company Now.)
Amazon says multicloud is at best a transition phase; eventually, enterprises will see the advantage of going all-in on a single cloud provider, to take full advantage of that provider's price-performance and special capabilities. That provider, according to Amazon, will of course be Amazon. Oracle has a similar line.
This alignment makes sense for Amazon from a sheer market share perspective. AWS is the overwhelming market share leader -- with more market share than all of its major competitors combined -- so of course they're going to encourage customers to pick a single cloud platform. (See Microsoft Growing Explosively, but Amazon Retains Huge Cloud Lead.)
And Oracle supports a single-cloud strategy because, well, they're Oracle.
Everybody else is trying to grab market share from AWS, so it makes sense for them to tell potential customers that the customers don't have to move all their business from AWS. Just some of it. Give it a try.
Which approach will win out? My crystal ball is broken on this one; both the single-cloud and multicloud guys make good arguments.
And each side is hedging its bets -- Amazon supports Kubernetes for multicloud portability () and Microsoft warns about the perils of going too far in avoiding cloud lock-in while also supporting Cloud Foundry and other multicloud standards (See Microsoft's Russinovich: Avoiding Cloud Lock-In Is Risky Too, Amazon Launches Managed Kubernetes Service, Amazon Joins Cloud Native Computing Foundation: Why You Should Care and Microsoft Goes for Cloud Foundry Gold.)
In other news: If you want something different from the typical Hilton vs. Marriott experience, try Airbnb. You might end up somewhere like this: I Skipped Google's Cloud Conference to Ride the Hotel Elevator All Day.
- Amazon Launches Managed Kubernetes Service
- VMware & Amazon Grow Hybrid Tie-Up to 'Very Large Scale'
- Microsoft Buying Deis to Boost Containers & Kubernetes
- VMware Lets You Fiddle With Your Hybrid Cloud to Cut Costs
- Microsoft's Cloudyn Buy Boosts Multi-Cloud Strategy
- VMware & Google Hook Up Kubernetes for Enterprise
- SAP's Prime Directive: Aggressively Cloud-Agnostic Strategy
- Oracle's Ellison: Amazon & SAP Use Our Database Because We're Better
— Mitch Wagner Editor, Enterprise Cloud News