ra2 studio - Fotolia
Angel Diaz, IBM's vice president of cloud technology and architecture, and his team are responsible for the technology, architecture and strategy behind IBM's hybrid cloud efforts. He is also IBM's leader for open technology. He recently spoke with SearchCloudApplications about the multicloud environment that developers face every day.
Today's multicloud environment is vastly different from where we were a few years ago. What is your view of this technology evolution?
Angel Diaz: As a computer scientist, I've been amazed at the evolution of technology, from the early days of assembler language to functional programming languages to object-oriented computing to the internet to the advent of service-oriented architecture -- business information put in an accessible way -- to where we are now with cloud and artificial intelligence. There has been a democratization of technology at every one of these steps. I couldn't predict back in the 1990s when I joined IBM Research that this would happen.
Is that evolution making life easier or more difficult for application developers?
Diaz: When you look back, it's easy to see the pattern. Over the years, the number of books that developers need to read in order to learn and be productive is being reduced dramatically.
As a developer, you no longer need to write a speech-to-text algorithm. Someone else has done that, and you call it through an API. What has happened in the multicloud environment era is that developers have fewer things they need to worry about in order to do innovative and productive things.
The challenge of that multicloud environment is there are lots of jigsaw puzzle pieces to assemble into a finished picture.
Diaz: When we look at the cloud market, cognitive computing and what's happening with artificial intelligence, there has been a shift from focusing on compute, storage, networks and base service. Developers know they have those, know they have data centers spread across 50 different geographies, know they have optimized networks and know they have a platform with containers that they can leverage. They can be better, faster and more valuable, and that is what gets business leaders excited. You can go from concept to creation of a new disruptive technology, and developers can build it iteratively and much quicker.
As the multicloud environment continues to expand, what are you seeing in new technologies that follow?
Diaz: When you look at the 20,000 people who attend our IBM InterConnect conference, you can see a democratization of technology and the driving of innovation -- whether it's the evolution of the cloud platform, the focus on business value, the evolution of data that developers can easily access and leverage or the ability to get at higher-order services.
Look at blockchain technology. When we started blockchain, we were thinking about how it could redefine the meaning of a transaction. Think about it: Transactions can now be peer to peer. There can be voting mechanisms. You can have automation around transactions without involving the traditional checks and balances, because they are all integrated into the intelligence of the network.
Do you see blockchain continuing to expand in presence and grow in importance?
Diaz: Yes. What we didn't know is that blockchain technology goes far beyond just finances. IBM just announced a blockchain-based carbon credit management platform in China that's based on the Linux Foundation's Hyperledger Fabric.
Now, you can get carbon credits and have companies negotiate how much carbon emissions they're allowed as long as the aggregate stays under a preset cap. That's an amazing use of technology that focuses on the value, where you don't have to worry about data centers or networking, because that's already standing up in the cloud.
Angel Diazvice president of cloud technology and architecture, IBM
Another technology that is growing in importance is cognitive computing.
Diaz: With cognitive computing, applications can analyze images or process speech. Once you understand speech, you can go further and do sentiment analysis on that speech. We can tell if you are angry, like when you're on the phone and start yelling, 'Operator.' If I'm on the phone with my airline, and the system knows I'm an executive platinum-level frequent flyer and I'm angry, it can take appropriate action.
When you go beyond sentiment analysis, data and cognitive computing start to merge. We can take unstructured data and apply structure and intelligence around it to do higher-order tasks.
No discussion of a multicloud environment is complete without discussing APIs. What's new that application developers need to know?
Diaz: We do a lot of things in an industry-specific way. We have an industry partner talking about a more contextual application-specific type of API. For example, in the financial services world, we are seeing the growth of fintech APIs. You have financial institutions that deal with payments and settlements. And, now, you have standards organizations that are defining how computers should interact with these systems. That's one industry-specific example.
We combined that with analytics and data capabilities, so developers can create new things. In the fintech world alone, we believe there are more than 100,000 developers. These developers can get the financial services interface into their own systems and build new, innovative cloud and mobile workflows around that. It opens up the business process, and in traditional financial institutions, it allows for an ecosystem of startups and new innovations to be built around their legacy core systems.
Applications are nothing without data. Where is the trend of moving data from legacy on-premises infrastructures into the cloud?
Angel Diazvice president of cloud technology and architecture, IBM
Diaz: It is the hybrid cloud that connects the two sides of the [infrastructure] equation with its integration capability. Think about this: More than 70% of the world's data still lies behind firewalls, in a private cloud or on premises. There is data gravity there -- and, sometimes, for very good reasons. It's foolish to think that you can pick up and move all that data or move all of your applications to the public cloud on Day 1.
We live in a multicloud environment. That could be private or public cloud, or even different [software as a service] vendors. What's needed is to get all of these different clouds to behave as one. We're doing that through standards -- OpenStack, Cloud Foundry, Kubernetes. What we've tried to do is make it easy for application developers.
On a mainframe -- and that's not a bad word -- you can run Linux and scale a hundred thousand containers in half a microsecond. Through an API connect capability, a developer can build a cloud-native application that securely and with scalability connects to back-end data. You can put a Mongo NoSQL interface onto that, build a mobile app and have a developer who knows nothing about mainframes access all of that.
How do you manage a multicloud environment?
Time is right for deploying microservices in your multicloud environment
Kubernetes, Docker, multicloud orchestration are a potent trio