Welcome to Oracle Cloud Infrastructure Innovators, a series of articles featuring advice, insights, and fresh ideas from IT industry experts and Oracle cloud thought leaders.
Enterprises are turning in droves to hybrid cloud computing strategies, especially for testing and development, quality assurance, and DevOps activities. But before the majority of enterprises can move on to more advanced hybrid cloud use cases, they'll need to overcome some lingering challenges.
I recently sat down with Bob Quillin, Vice President of Developer Relations at Oracle Cloud Infrastructure, to discuss Oracle’s cloud native direction. We discussed the biggest trends in hybrid cloud computing and the major obstacles—things like skills shortages, resistance to cultural change, and rapidly evolving technologies—that often stand in the way of adoption.
Listen to our entire conversation, followed by a condensed written version:
One of the common trends I'm seeing in just about every enterprise is the move toward hybrid cloud strategies. What are you seeing on that front?
Bob Quillin: We're seeing a lot of demand and interest in hybrid cloud. People have been trying out different models and patterns and testing out different technologies, and there have been some challenges. But one of the first major areas where we're seeing a lot of traction with customers is using the cloud for development, quality assurance, DevOps, and for running tests, with production still largely on premises. Many people feel more comfortable in their on-premises environment for certain production applications. But with a cloud native and DevOps environment in the cloud, you can spin up, spin down, and support a variety of testing, staging, and QA projects. It gives you a lot of elasticity, it's cheaper, and the test cases can run in containers. Sometimes people say, "Oh, I can't run my database applications in the cloud." Well, that isn’t the case for test and QA use cases. You can put them in a container, run the test, break it back down, and you're good to go. The disposability and quick reusability of these environments is where we're seeing a lot of success, and that is where a lot of people get started.
What's the next step on the road to a hybrid cloud strategy?
Quillin: The next step is getting to the point where you have a platform that gives you confidence that you can develop on the cloud or on premises, and that you have bi-directional portability—on premises to cloud or cloud to on premises. Ensuring that kind of application portability is the next major pattern we've seen. Disaster recovery and high availability deployments is the third approach we've seen. For example, people will mirror their application in the cloud to have it available. But they keep running the existing application on premises so they can failover if they have a disaster event. Disaster recovery is one of the classic hybrid models. Those three areas are the ones we see being most successful right now.
Are organizations using hybrid strategies at all in more advanced areas?
Quillin: There are two more use cases we're seeing that are more advanced. One is a workload balancing application that's able to run both on premises and in the cloud. It lets you choose where to run each workload based on its regulatory requirements, governance, latencies, whether it’s a new or legacy workload, etc. This approach requires a bit more sophistication and a little more targeting.
The other big one that people have been working toward for a long time has been cloud bursting, where users can expand resources into the cloud dynamically back and forth. Or, users enlist some kind of federated automation where, based on performance or quality of service, I'm able to choose where I run my application and have a federated, single view of all that. These use cases have been highly desirable from an enterprise perspective. But what's been lacking is a platform from which to do it and a framework that enables it.
Let's talk a bit more about challenges. I'm sure almost every organization that you deal with is facing certain setup challenges in deploying, particularly to the hybrid model. What are you seeing?
Quillin: Cultural change and training continue to be inhibitors—and I think those roll up into an overall operational readiness challenge. Organizations are struggling with how to get started on this. At Oracle, what we're providing is an easier way to get started. The Oracle Cloud Native Framework provides a set of patterns and a model that gives the customer a supported blueprint for hybrid cloud. The next challenge is dealing with portability complexities related to a variety of underlying integration issues, including storage, networking, and the wide variety of Kubernetes settings and configurations. A related challenge—and this is one of the dark secrets of cloud native—is that there are a lot of “devil in the details” problems based on the rapid rate of change of Kubernetes, its quick release cadence, shifting APIs, and the general way the technology is rocketing forward. What you need is a vendor that supports you through these changes by supporting a bi-directional portability model. At Oracle Cloud Infrastructure, we're helping organizations through this process—and we're not going to leave them high and dry by using a proprietary approach. We're committed to open standards.
Many organizations think that open source is great. But there are also those who think that sourcing software from a single proprietary vendor can be cheaper due to the DevOps and maintenance costs associated with open source. What are your thoughts on that?
Quillin: All sorts of studies have been made around organizations that use an open source and DevOps culture, and they're always faster and more successful in terms of business agility. But also, the developers are happy. It's true that some on the business side of an organization would choose proprietary technology. But if you really want to recruit the best developers, you're going to want to work in open source because that's the most marketable set of skills today. You get happier developers, you can recruit better, and you get the best development teams.
Oracle is a platinum member of the CNCF (Cloud Native Computing Foundation). How does the CNCF help in terms of enabling enterprises to overcome these challenges?
Quillin: I think the most important thing they've done—which is amazing to me—is they've enabled the market by creating a standard cloud native platform based around Kubernetes. That's been their crowning achievement so far. If you remember back to just a few years ago, everyone had their own orchestration technology and it was all over the place. That's settled down now. The CNCF has created stability and enablement for the market.
What is next for Oracle and CNCF?
Quillin: The challenge is to continue that success. There's some next-level tooling that needs to come out. Some of the fastest-growing projects in the CNCF are around monitoring and tracing and logging, around networking and storage, and around the best ways to manage a Kubernetes environment and connect it to existing storage and networking infrastructure. Kubernetes is growing, but what's really growing faster, which is a good sign, are the things that make Kubernetes more manageable, more secure, and more integrated into your existing infrastructure.