I’ve just co-authored a new book on OpenStack architecture. Part of my motivation for writing it was to help people avoid the most common mistakes when embarking on a new OpenStack initiative. First, I’ll talk about avoiding three of these errors, and then I’ll talk a bit about the new OpenStack book.
Mistake 1: Failing to get enterprise-wide buy-in
Switching to cloud architecture is not a simple shift in technology strategy: it’s a transformative move in which everybody across the enterprise must believe and commit. If you don’t win this buy-in before you make your cloud move, you won’t be able to leverage all the technology’s potential.
Moving to a cloud architecture from legacy bare metal or virtualization platforms won’t do much good if you don’t fully utilize it through your business processes and developer culture. OpenStack clouds are not “build-it-and-they-will-come” technologies.
The best approach is to use simple, basic projects to get small wins for agile and cloud development. Then, the people in those projects can become cheerleaders for bigger, more powerful OpenStack clouds.
If you don’t try the slow and gradual approach, you run the risk of accelerating the rejection of OpenStack. Even with executive sponsorship, rushing into OpenStack can get you a great platform with no tenants.
Mistake 2: Taking on too much
OpenStack has an impressive (and beguiling) feature set. It’s human nature to want to use all the bells and whistles from the get-go; however, OpenStack has so many impressive capabilities that trying to use too many of them can be frustrating and overwhelming.
Here’s what’s likely to happen if you’re new to this: you’ll see what you want your cloud to look like, then get confused by all the architectural options in front of you. Of course, you should have some familiarity with cloud architecture before diving into OpenStack, but learning how all the interoperability works is a challenge for beginners.
My co-author and I hope to bridge that gap with our new book, “Open Stack for Architects.”
Mistake 3: Lack of planning
OpenStack deployments must be planned properly. Many new cloud architects go astray in the planning department.
I’ve seen too many enterprises with weird-sized flavors of the technology installed on the same infrastructure. This produces inequal, orphan resources that don’t get consumed. This isn’t a big problem on a small scale, but at enterprise scale it can be extremely costly.
The point of using OpenStack in a private cloud is to maintain an optimum balance, where your capital expenditure for in-house data center technologies is more economical than the cost of moving all your cores to the public cloud. Big companies with tens of thousands of cores can recover thousands of dollars in wasted cap-ex simply by doing a better job of planning their OpenStack implementation.
“Open Stack for Architects” provides a roadmap for this kind of planning.
Find out more in ‘OpenStack for Architects’
I co-wrote “OpenStack for Architects” with Red Hat’s Chief Architect Michael Solberg, one of Red Hat’s top OpenStack architects. The book will help you:
- Become familiar with the key OpenStack components.
- Set up a lab to design and build progressively complex OpenStack deployments.
- Write effective documentation for your company’s architecture teams.
- Apply agile configuration management techniques.
- Integrate your identity management, provisioning and billing systems into OpenStack.
- Set up a powerful virtual environment where users can interact.
- Implement enterprise security guidelines for your OpenStack deployment.
- Create a product blueprint that quickly delivers features and functions to platform users.
Who does this book help most?
If you’re a cloud architect that is planning and leading an OpenStack cloud implementation, “OpenStack for Architects” is a great place to start. You’ll get answers to key deployment questions like:
- How does it differ from traditional virtualization?
- How do I choose hardware or third-party software plugins?
- How do I integrate the cloud into my existing infrastructure?
While the book is more focused on the planning and architecture phases of cloud deployment, you’ll get some nuts and bolts, too — a number of code examples and integration patterns are included.
We’ll guide you through all the major decision points of designing an OpenStack private cloud. Our advice draws on our own experiences from leading successful OpenStack projects in multiple industries. Every chapter has lab material to help you install and configure production-class OpenStack clouds.
The whole idea behind our book is to ensure that your OpenStack initiative meets your organization’s needs, which in turn guarantees a successful rollout.