Once again, the OpenStack Summit is nigh and this time it’ll be in Barcelona.

The OpenStack Summit event is an opportunity for Operators, Developers and Users alike to gather, discuss and learn about OpenStack.

What we know is that there’s going to be keynotes, design sessions for developers to hack on things and operator sessions for discussing and exchanging around the challenges of operating OpenStack. We also know there’s going to be a bunch of presentations on a wide range of topics from the OpenStack community.

What we don’t know, though, is which presentations are going to make the schedule. It’s complicated when there’s more than a thousand submitted entries.

Presentations are first submitted in a category (or ‘track’), such as “How To & Best Practices” or “Upstream Development” and then the community is invited to vote on which presentations they’d like to see. After the vote is concluded, elected volounteer track chairs takes the vote into account as one of the factors to choose which presentations make the cut.

I’ve already done my fair share of voting for presentations I’d like to see and thought it’d be a good idea to share some of the presentations I would be attending.

Also, for the first time ever, I’ve participated in submitting presentations to try and share the great work we’ve been doing to package, test and ship RDO. So, of course, I want to advertise these talks so that we can have the opportunity to highlight what makes RDO special.

My presentations submissions

My three presentations can be found on the voting website by searching for “David Moreau Simard”.

From Our Code To Your Production Cloud

With my colleagues Emilien Macchi and Pradeep Kilambi from Red Hat

For an OpenStack developer, it’s sometimes challenging to understand how a feature will be integrated in production. And sometimes cloud users might wonder how a feature has been developed, integrated, and automated in a production cloud deployment.

In this talk, we’ll laugh about the famous “it worked on devstack”. We’ll also show you step-by-step how we develop a feature in OpenStack and then integrate it in TripleO, which is a project aimed at installing, upgrading, and operating OpenStack clouds.

Come and see how your production cloud is developed and deployed under the hood.

Behind the scenes of testing RDO, an open source community OpenStack distribution

With my colleague John Trowbridge from Red Hat

OpenStack strives to provide APIs that abstract complex operations in order to mix and match different OpenStack projects, hardware, and network topologies. We verify that aspiration by testing an ever-growing matrix of possible combinations of installing and configuring OpenStack. This process is difficult and intensive.

In this presentation, two engineers of the RDO community will share their experience testing trunk across these different installation scenarios.

Join us to dive into the projects the RDO community uses to tackle this challenge, and learn how you can get involved.

Anatomy of an OpenStack distro

With my colleague Javier Pena from Red Hat

The OpenStack development process rarely finishes when code is tagged for a release. Throughout the life cycle of a release, OpenStack distributions package, maintain, test, and turn OpenStack into a product that’s easy to consume by a large variety of users on their favorite operating system.

In this presentation, you’ll get an inside look at the work that goes into shipping a stable OpenStack distribution on the same day as the upstream release. We will explain how the RDO community distribution continuously packages OpenStack from source, how we test it and how we ship it. We will share our experience, our challenges and how we overcome them in collaboration with the OpenStack community.

Presentations I’d like to attend

These are some of the presentations that, if they are selected, I’ll try to attend.

Zuul v3: OpenStack and Ansible Native CI/CD

This one will give insight as to the future of how the future will look like for the future of testing OpenStack. The OpenStack Infrastructure team recently replaced Jenkins with Zuul “v2.5” and they are moving full steam ahead towards Zuul v3 with many great improvements on the roadmap.

By James Blair

The OpenStack Project Infrastructure team developed Zuul to manage its project gating system. Zuul is a flexible, general purpose system to integrate code review and automated testing. The team is preparing for the next major version of Zuul which will provide a number of new features to developers:

• Native Ansible job definition
• Drive testing and deployment from the same playbooks
• In-repo job definitions
• External cross-project dependencies
• Flexible multi-node testing

This presentation will give an overview of how Zuul works and how OpenStack developers can use these new features in their project’s gate.

Using Kibana & ElasticSearch with OpenStack

I’d like to hear more around logging. Logging is an important part of operating OpenStack at scale. Troubleshooting an issue on an “all-in-one” OpenStack deployment is sometimes hard enough. Trying to pinpoint a problem when you have hundreds of servers is really something else.

By Lida He and Nebu Mathew

This presentation will provide an overview of Kibana and ElasticSearch. We will cover how they can be used in a distributed system and integrated with OpenStack. We will go into detail on how ElasticSearch shards logs and applies algorithm for search. We will demonstrate the capabilities of the Kibana user interface - searching logs, trending issues, exporting log files and more. Finally, we will provide examples of cases where Kibana was used to analyze and solve real world issues in OpenStack.

Best practices for deploying & integrating OpenStack and Ceph with Ansible

I love Ansible and I’ve never got to use OpenStack-Ansible to deploy OpenStack before. There’s some really great people working on the project and I’d like to have a look at how they do certain things - this talk looks like it’ll give a good overview.

By Syed Armani

This session will provide a detailed overview of deploying OpenStack and Ceph with Ansible. We will go through the best practices for deploying OpenStack with openstack-ansible and Ceph with ceph-ansible. What we will cover:

• Brief overview of Ansible, OpenStack and Ceph
• How to install OpenStack with openstack-ansible and Ceph with ceph-ansible
• How to integrate OpenStack with Ceph
• How to enable Swift API compatible access to your Ceph with radosgw
• How to manage Galera cluster failure
• How to manage RabbitMQ cluster partitions
• How to add/remove controller nodes
• How to add/remove compute nodes
• How to leverage dynamic inventory
• How to destroy and rebuild service specific containers
• Upgrading your OpenStack environment
• How to make environment specific configurations
• Provide tips and tricks for daily operations

Downstream Gate - How to set up and run pre-merge, gating CI using OpenStack CI tooling

Yup. Testing OpenStack is hard, even downstream once everything is “supposed” to work. I’d love to see how others are tackling this challenge.

By Dan Moravec, John Warren and Clint Byrum

In this session, we’ll show how to set up the OpenStack CI tooling to run tests “downstream”, i.e., after code has merged upstream. Everyone is talking about CI/CD pipelines these days. But “CI/CD” means a lot of different things to different people. We’ll demonstrate techniques that bring the powerful OpenStack continuous integration (CI) tooling to benefit Cloud Operators. Learn how these tools allow you to run CI to test your specific deployment with the exact code you plan to deploy to production….before the proposed changes to deployment tooling or local patches are even merged into your local code repository. Once you are able to run CI in this manner, continuous deployment (CD) can be achieved with a much higher confidence level.

OpenStack on Kubernetes - Lessons learned

I’ve never had the chance to try Kubernetes out yet. This looks like a good opportunity to learn about it and see what problems people have encountered with it in the context of OpenStack.

By Sergey Lukjanov and Piotr Prokop

Due to its important number of components, deploying and managing OpenStack is known to be a complex and error-prone process.  On the other hand, containers recently introduced consistent deployment capabilities and container management systems brought robust application lifecycle administration at scale. Therefore, we’ve decided earlier this year to bring the worlds of VMs and containers together by containerizing OpenStack services and managing them using Kubernetes. This effort enables companies to deploy a scalable and self-healing OpenStack platform in less than 5 minutes on top of Kubernetes. During this session, we will share the experience that we gained during the kubification of OpenStack, demonstrate how this work simplifies and accelerate OpenStack deployments and operations, and describe how what was initially a proof of concept is growing into a production-ready product.