The second edition of “Mastering OpenStack” offers expert tips on tweaking your deployments.


A pair of global OpenStack virtuosos have paired up again to help take your OpenStack deployment to the next level.

Chandan Dutta Chowdhury and Omar Khedher are the authors behind the second edition of “Mastering OpenStack.” Based in India and the Netherlands, respectively, they talk to Superuser about keeping up with the increasingly fast landscape and working across time zones.

Aimed at intermediate- to advanced- level users, the latest edition of the 430-page manual aims to cover designing, deploying, and managing OpenStack-based clouds in mid-to-large IT infrastructures.

It will be released this month, but you can already pre-order the eBook version from Packt Publishing with a 30 percent discount (use code SUMO30) until Jan 15, 2017.


What are the most significant updates to this edition of the book?

Chowdhury: OpenStack has a lot of updates from release to release…There has been a considerable interest in container technology like Docker and Kubernetes. We have added chapters on network functions virtualization (NFV) and software-defined networking (SDN) which has seen significant contributions in OpenStack. We have also added more details on storage virtualization.
Khedher: The new edition offeres updates on best practices on deployment and administration on OpenStack in production, too.

Who will this help most?

Chowdhury: We assume readers already know the basics. People deploying OpenStack at scale in their organizations can find details on deploying and operating a large scale OpenStack cluster.
The book also provides solutions for high availability and reliability needs of a production cluster and a guide for sizing your OpenStack installation based on your deployment needs.

What are the most common mistakes people make?

Chowdhury: People think automation will solve all their deployment problems in OpenStack. Yes, automation helps in deploying OpenStack but it is not a replacement of skills needed to deploy and operate an OpenStack production environment. The knowledge of the underlying design is invaluable in keeping your OpenStack infrastructure running.
Other points often overlooked while moving to a production environment is planning for the redundancy of services, anticipation of service failure, software upgrades and planned maintenance.  You need to plan for redundancy in OpenStack clusters.
Khedher: OpenStack is great solution to build your own private cloud computing environment but, people should see it also from architectural perspective and understand how the basic components interact…Users should be confident enough in operating the base components efficiently. This can raise a challenging task on how to automate the infrastructure deployment in an effective way… Additionally, users should look further ahead and plan for growth, although the deployment project for example may start small…

How would you define an OpenStack Master?

Chowdhury: As OpenStack is an ever evolving project and it encompasses a lots of project under a single umbrella it is difficult for a person to master all of these projects. That said there is still the core set of projects like the compute, network and storage that are the essence of any OpenStack installation. So according to me a master OpenStacker is one who holds significant mastery on the core OpenStack projects and at the same he is well aware of width of newer projects that are added to OpenStack at each release.
Khedher: OpenStack Master…This is very relative! Every OpenStack project is a wide area where people are fully dedicated to join and contribute to one project. I might assume that the mastery of OpenStack technology is based on a good understanding of the basic components of this platform and having a conceptual background on how to make from them a consistent, automated and effective infrastructure.

There’s so much you can learn online — why buy a physical book (or even an e-book)?

Khedher: OpenStack official documentation  is a great place to learn. Relying on online resources can be fruitful to try things in a test environment and have proof-of-concept up and running. On the other hand, books are not only about writing code and encouraging readers to try them out, but a well-structured workflow and topics to guide the reader to real production…”Mastering OpenStack” is not just about ‘how to’ or ‘do, watch and see,’ but about sharing experiences that motivate readers to take an advantage of them and use them as a pattern on their new journey.

Collaboration is a big part of OpenStack. You wrote this on two different continents – what tips do you have for working on deadline across time zones?

Chowdhury: The most important aspect of collaboration and community based development is to plan for the deliverables and attaching responsibilities to the fellow contributors. The other aspect of succeeding in a community is to learn from others, remain open to suggestion and constructive criticism.
Khedher: Deadlines are sometimes overloading the energy of the author to deliver a high quality content on time. Following up suggestions and recommendations from the co-author is very constructive and may guide the writing workflow to better way and move faster.