CentOS Connect

Upgrading between major versions of CentOS Linux was always a pain, and required tons of manual labor. Project ELevate is an open source upgrade tool that allows you to move between major versions of many of the most popular RHEL-equivalent operating systems without the tedium of moving your content! Come talk to the maintainers of this project, learn how you can use it in your environment, and discuss ideas for improving it!
https://almalinux.org/elevate/

This will be a combined meetup led by the Hyperscale and Alternative Images SIGs. This meetup will be an opportunity to sync up and collaborate on the next year of work within both SIGs.

Of particular focus will be the bringup of support for CentOS Stream 10 based deliverables (Hyperscale Spin, alternative desktop images, cloud and ARM disk images, etc.) and future efforts around growing the CentOS community.

Join us for the Rocky Linux - Technical Steering Meetup at CentOS Connect in February. This 2-hour session will bring together members of the Rocky Linux teams and community to outline and discuss the roadmap for 2025. We'll delve into key development milestones, upcoming enhancements, and strategic goals for the Rocky Linux project. This is also an open platform for attendees to ask questions and gain insights into getting started with Rocky Linux. Whether you're a seasoned contributor or new enthusiast, your input and curiosity are welcome. Come be part of the dialogue shaping the future of Rocky Linux!

This meetup will be to talk about all things OKD and Kubernetes. Discussion will be on all aspects of the effort. Feel free to bring your questions, comments, things you would like to work on. Everyone is welcome to attend.

CentOS Stream is a Linux distribution built by Red Hat Enterprise Linux (RHEL) engineers as part of RHEL development.

Innovation within the OS happens in Fedora. CentOS Stream provides a solid base for innovation on top of an OS. Many CentOS Special Interest Groups (SIGs) are building on and extending it in all sorts of interesting ways without needing to reinvent the wheel.

CentOS Stream also serves as a contribution path to RHEL itself. The source, the builds, the release, it all happens in the open.

Come and learn about what's new, what's coming, how to get involved both directly and through CentOS SIGs.

This session highlights the OKD community’s initiative to evolve Kubernetes scheduling into a dynamic, collaborative framework. Building on the experience gained from developing the Multiarch Tuning Operator for OpenShift, the initiative leverages Kubernetes' scheduling gates mechanism to go beyond multi-architecture-aware scheduling. It introduces a framework where multiple controllers can compete to inject augmented information into pod specifications, enabling smarter and more efficient pod placement across Kubernetes clusters.

Rooted in OKD, the upstream distribution of OpenShift, and running on CentOS Stream CoreOS, this effort showcases how the OKD community is driving innovation for scheduling in Kubernetes Clusters. The framework optimizes workload placement while coordinating with descheduling and autoscaling components.

This talk will explore how OKD’s community-driven approach connects observability platforms with Kubernetes’ scheduling ecosystem, closing the feedback loop for improved performance, SLA guarantees, cost savings, and energy efficiency. Attendees will also learn how this framework lays the foundation for a fully distributed, intelligent placement system for Kubernetes workloads.

Join us to discover how the OKD community is extending its vision through CentOS Stream CoreOS, fostering collaboration and innovation to advance Kubernetes scheduling.

After a long time working on it, we finally launched the new website and docs site with the CentOS Stream 10 announcement. There was a lot of frantic scrambling among the Artwork, Infra, Docs, and Promo SIGs. There were snags. There were delays. But we finally got it out the door.

This talk will explore what we did. But more importantly, it will explore what we didn't do. We had to think hard about what work we could defer to meet deadlines. Learn how you can contribute tonight from the comfort of your hotel room.

We are currently going through a big shift in OKD. We are working to have everything entirely built on CentOS stream. This talk will explain what we want OKD to look like and the steps we are taking to get there,

At RDO, we are experimenting building the OpenStack services from source instead of packaging them first with RPM. This initiative is conducted alongside the Konflux effort which is taking place within the Fedora community.
We'll present a PoC of an Openstack service built from source in the Cloud SiG Konflux tenant.

On the dates of CentOS Connect 2025, we get to celebrate exactly 4 years since the release of the very first beta version of AlmaLinux.
While being RHEL (and later CentOS Stream) derivative AlmaLinux still does a lot of things differently both on the distribution and tooling sides.
Our build system and mirror service are just the tip of the iceberg. In my talk I'd like to focus on more things we do differently, like:
- How we build images
- How we produce errata
- How we do OpenSCAP profiles and OVAL data
- How we support additional hardware and older CPUs
- How we extend virtualization support

... and many more.
This can be useful for users and developers to look at familiar features and processes from a new angle.

What if detecting bugs and vulnerabilities in RPM-based distributions could be seamless and fully automated?

OpenScanHub is a service for static and dynamic code analysis. It was internally used inside Red Hat to scan releases of RHEL for more than a decade and was open-sourced in 2023.

OpenScanHub can fully automatically scan RPMs and has the ability to do differential scans that helps in finding bugs that may be introduced on package updates and new distribution releases. By default, it supports static analyzers embedded in GCC, Cppcheck, ShellCheck, find-unicode-control, Clippy and is extensible to support other analyzers. It can collect reports from various analyzers at a single place to make it easy to analyze them.

OpenScanHub was recently integrated with Packit, a CI/CD solution for automating RPM package builds, tests, and distribution releases. This new integration performs differential scans on pull requests, so potential bugs may be found during the pull request review process and would not be introduced into the codebase.

In this talk, we will share ideas about how CentOS Stream and its derivatives may benefit from OpenScanHub.

The CentOS Infra Special Interest Group is there to serve the whole CentOS Ecosystem, especially the other SIGs.
What has been achieved during the 2024 year ? What are some other goals for 2025 ? Let's present these though slides but also Q&A (hearing from SIGs themselves !)

This presentation will provide an update on what the CentOS Hyperscale SIG has been working on, what work has been done by the Hyperscale SIG in CentOS Stream, what deliverables are available, how to use them, and what's coming up next.

CentOS Hyperscale is constructed with the combination of CentOS Stream, Fedora EPEL, and our own produced packages. This gives us a broad content set, but since each of these are released with their own cadences, it becomes important to create discrete collections of this for various purposes (notably integration testing).

This talk will discuss the problem and share the solution created for the Hyperscale SIG, and show how other CentOS SIGs and communities can benefit from it.

CentOS Alternative Images SIG has progressed alot in the past year. Troy will go over all the new Images we adding this past year and what we have planned next. He will also do a demo some of his favorite images.

OpenHPC is Linux Foundation project which tries to provide an easy starting point into High Performance Computing (HPC). Currently the OpenHPC projects supports Leap 15.5, openEuler 22.03 and different RHEL 9 clones (AlmaLinux and Rocky Linux). For those distributions OpenHPC provides RPMs and validated recipes which guide the user to a running HPC cluster.

In this session I want to give an introduction why OpenHPC exists and what special requirements HPC systems have, how OpenHPC builds its RPMs and how OpenHPC validates its released recipes with hundreds of tests for each release.

For some time, Packit’s main target had been Fedora. But we have something for the CentOS Stream community as well. Specifically for CentOS SIGs this time.

Providing builds and CI for your SIG is not easy, and with Packit, we thought we could be of help. We were asked about this a long long time ago, but last year, Christian Glombek sent us the first contribution that kicked off the actual work and together with the Packit team, the work on automation for CBS Koji builds started for real. Just another Koji instance one would say. We’ve come a long way since then and learned our lesson. Come and see what it takes to automate RPM builds on CBS Koji in reality and how you can benefit from our work.

During the talk, we’ll show what we’ve managed to finish and what are our plans for the future.

Packit can be a natural choice when it comes to Fedora automation. But did you know that the very same works for EPEL?

Let’s take a look at how Packit can help you get a new package version from the upstream release to the user. We’ll take a look at multiple approaches to getting your new version to dist-git, and we’ll continue with Koji builds (no worries, side-tags are supported) and we’ll finish with Bodhi updates.

Let’s meet during lunch to see all this at work and help us help you save some time when maintaining a package.

Extra Packages for Enterprise Linux (EPEL) is a yum repository of community maintained packages for use on CentOS Stream and Red Hat Enterprise Linux (RHEL). For most of its history, each version of EPEL was made available after the corresponding major version of RHEL. This slowed down package availability, which then slowed down adoption of new RHEL major versions. In EPEL 9, package maintainers were able to build against CentOS Stream 9 early to have a large number of packages ready before the RHEL 9.0 launch.

For EPEL 10, the EPEL Steering Committee is expanding that strategy to all minor versions of RHEL 10. This will improve support for CentOS Stream and for specific minor versions of RHEL, resolving several key pain points of users and maintainers. Attend this talk to learn more about this bold initiative and the results achieved so far.

Foreman is a robust, open-source solution for provisioning and managing CentOS systems at scale. This talk will highlight how Foreman simplifies the provisioning process for CentOS environments using PXE-based booting, image-based workflows, and integrations with hypervisors like Libvirt and VMware.

We’ll also explore recent advancements such as Secure Boot and IPv6 support, ensuring that Foreman remains compatible with modern CentOS infrastructure needs. A live demo will demonstrate how to efficiently provision CentOS systems, helping attendees streamline their workflows and manage environments with confidence.

Key Takeaways:
- Learn how to provision CentOS systems efficiently with Foreman.
- Explore advanced features like Secure Boot and IPv6 for CentOS.
- Gain practical insights from a live provisioning demo.

Many, if not most, deployments of CentOS Stream and its downstreams (RHEL and derivatives) require packages from the Extra Packages for Enterprise Linux (EPEL) repositories, which requires bootstrapping for every major EL release. This is a follow-up to previous CentOS Connect talks, discussing how I am using poi-tracker to track packages, export them to ELN Extras workloads, and then use ebranch to branch and build these packages and their dependencies in EPEL 10. I plan to release the first stable versions of poi-tracker and ebranch at Connect, after stress-testing ebranch across two EL releases (9 and 10).

There’s an exciting potential for bootable containers, which allow you to build and manage a full operating system just like a container image, and recently, Red Hat announced it’s intention to donate the tool to the Cloud Native Computing Foundation (CNCF). However, for AI/ML workloads which require a complicated stack of dependencies, this technology helps curate the delivery of a full stack for training and inferencing, for example with Red Hat Enterprise Linux AI. Join us as we put together an operating system for running an AI-enabled application with CentOS Stream, using an InstructLab fine-tuned model from our local developer workstation. With bootable containers, our deployment workflow is simplified, with flexibility for dynamic requirements and environments in building the next generation of Linux workloads.

The Foreman and Pulp projects are upstream to Red Hat Satellite and together contain more than 750 SRPMs layered on top of Enterprise Linux. To keep this all maintained with a small team we need automation. See how we utilize COPR, Jenkins, GitHub Actions, CentOS CI, Ansible, gem2rpm, pyp2rpm and more to deliver both nightly and stable releases.