Virtual PulpCon 2025!

[ggainey]

Virtual PulpCon 2025!

Where?: Online!
When?: December 1-5, 2025, 0800-1200 GMT-5 (US ET)
CFP ends: Oct 31, 2025
CFP interest voting (if needed): Nov 3 - Nov 7, 2025

Virtual PulpCon is coming and we would like to welcome you and encourage to participate in any way you prefer, whether it’s presenting, active participation in discussions, or just listening. Similar to what we did for Virtual PulpCon 2024 , we will host half-day sessions, with presentations recorded and made available on our youtube channel .

To propose a session for PulpCon 2025, please add it to the list below, along with your name. Add an estimated timeframe (25 or 50 min) for the session. Each talk should leave at least 10 min at the end (5 for Q&A and 5 for a break).

This CFP ends on October 31st. (This is a relatively short CfP window - but the reality is that the vast majority of CfP responses come in either the first week of the window, or on the last few days.)

Some possible “themes” for talks:

• Lessons learned from production use of Pulp
  • Experiences running Pulp “at scale” (i.e., large datasets, large number of client-consumers)
  • What features do you use/not care about?
  • What do you workflows look like?
• Experience integrating with Pulp
  • If you could have a brand-new API - what would it look like?
• What do you add to Pulp?
  • What tools do you use to make Pulp fit into your use-case?
• What things would you wish Pulp did differently?
• “Lessons learned in the last year”
  • both from the Pulp team and Pulp users!

For proposals, please define your “intended audience”, preferably with a focus on either “developers” or “users”.

NOTE : It’s fine to add a “rough idea” for a talk below, to be refined as we get closer to PulpCon!

If you have any questions/comments, feel free to add those as nested bullet points too.

Note that we plan to leave some timeslots open for topics that come up during the week of PulpCon.

Talk Proposals:

Template:

[Focus] "Title" - Intended duration - Speaker's Name
    * short description

[iballou]

[Developers] "2025 Katello & Pulp Integration Updates" - 25 mins - TBD (maybe Ian)
    * Discuss what Pulp-related features have been built since PulpCon 2024.
    * Give a preview of current and future development plans for Katello, including our prep for the future Pulp 4 upgrade.
    * Give some details about how Pulp is working for our real users (unless this ends up as a full second talk)

I’d like us to give a second talk too perhaps diving into some specific feature - TBD.

[hyagi]

[Users] "pulp_python vulnerability report" - 25 mins - jobselko and hyagi
    * Introduce the vulnerability report for pulp_python
    * Give a demo of usage

[ggainey]

[DEV] "Content-labels in Pulp" - 25 min - ggainey
    * Describe the new content-labelling feature
    * discuss caveats
    * demo

[Geoffrey_Wilson]

[Users] "Introducing Pulp Manager" - 20m - Geoffrey Wilson (@suprjinx)
       * A quick introduction to Pulp Manager (recently open-sourced in the Pulp GitHub org)
       * Demo it   

[rfguimaraes]

[Users] "Small-Scale Pulp Deployment with Quadlets and NFS" - 25 mins - Ricardo Guimarães (@rfguimaraes)
    * Present a small-scale, multi-node Pulp deployment without Kubernetes.
    * Overview of the architecture and design decisions.
    * Discuss main challenges on the interaction between container runtimes, Linux file permissions and ownership, and Pulp itself.

[x9c4]

[DEV] "Pulp-UI" - 25 min - x9c4
    * Discussion: What do we have, where do we want to go?

[StopMoitonCuber]

[Users] "Operating Pulp on Kubernetes - a 3 year recap" - 25 Minutes - Ruben Simons
    * Present medium-scale Pulp deployment for internal use in HPC center
    * Present tech stack used around kubernetes deployment
    * Showcase tooling/automation around pulp (especially involving RPM building & distribution)
    * Discuss some of the challenges encountered within the operation

In FZ Jülich, one of the biggest HPC centers in Europe, we have been using Pulp for the last ~3 years. I wanted to share some of the journey within a talk, though I have not been too engaged with the community so far.

In case the proposal looks interesting I would be happy to share a bit on it. As the talk is not prepared yet, I could focus more onto specific details that are more interesting for the community.

[Samir_Jha]

[Users] "Flatpak Pre-Installs via Satellite and Capsule Registries" – 30 mins – Bruno Ciconelle (Anaconda) & Samir Jha (@sjha4) 
* Demonstrate how Pulp-backed flatpak registry in Satellite and Capsules enable Flatpak mirroring and delivery.
* Show Anaconda integration for pre-installing Flatpaks during provisioning using Sat/capsule registries.

[Odilhao]

[Dev] "Building Pulp Container for Katello" - 25 min - Odilon Sousa
    *  Containers on Foreman/Katello are finally here.
    *  It's time to discuss how we will do the machinery to build our own images.
    *  it's buildah? It's Konflux? 

[gerrod]

[Users] "Exploration of the new PEPs pulp_python now supports and how to use them" - 30 min - Gerrod Ubben
* Going over all the new features related to PEPs that have been added to pulp_python in the last year
* Showing off the new workflows that take advantage of these new capabilities
* Preview of future PEP support

[Dev] "Building Pulp images: Gotta go fast" - 30 min - Gerrod Ubben
* Talking about the improvements to the pulp-oci-images CI that have been made recently
* Discuss the tradeoffs we face with our limited resources
* Share some ideas for even more potential future improvements

Some follow up advertisements for new pulp_python features, vulnerability reporting isn’t the only one! Also, more info on how to build Pulp. You shall learn how to build your own!

[bmbouter]

This is actually a talk from @dkliban

[Users] This talk will examine some of the challenges I encountered while helping various teams at Red Hat integrate their CI/CD pipelines with Pulp. The solutions to these problems required changes in client tooling, Pulp’s REST API, and the Tasking System architecture. As we examine each problem we will take a look at Bash and Python scripts that help upload and organize content that is produced by multiple different build systems at Red Hat. I will also highlight some of the new features that were inspired by the problems we experienced over the last year.

[bmbouter]

[Users]

After another year of running Pulp as a Service at scale, we’ve learned hard lessons about what it takes to operate a multi-tenant content management platform.
This talk shares the technical challenges we faced and how we solved them:

Key Takeaways:

• Scaling the database - Loading PostgreSQL at scale, optimizing DB queries, and connection pooling.
• Upload performance - Handling synchronous uploads at high volume, dealing with upload timeouts, and architectural improvements.
• Authentication integration - Working with external auth systems you don’t control and the architectural patterns that saved us
• OTEL monitoring at scale - Implementing OpenTelemetry observability across a distributed system serving production workloads
• Self-service everything - Moving from manual operations to tenant self-service and what that means for API design and automation

[dralley]

[Users] Pulp 4 - Plans, Status and Current Issues - 25 min - Daniel Alley

• Talk about what the Pulp project plans for Pulp 4
• Talk about where it’s at, timelines, blockers

[pedro-psb]

[DEV] “Tasking System work in 2025” - 25 min - Pedro Brochado
* Introduction of the tasking system implementation
* Share the problem of scaling the workers that Pulp Services is having
* Discuss some accomplishments, ideas and limitations

[anthomas]

[Users] Pulp Roadmap - 25 mins - Andrew Thomas

• Looking back on our work this year
• Discuss our plans/milestones for next year

[quba42]

@iballou, so I read in several places you are collecting questions on interesting use cases for Pulp within Katello/Foreman. My question is how you feel about the following use case: Let’s say I have a Foreman plugin (not necessarily Katello). And I want this plugin to have access to some standardized content to provide its functionality. So, for example, I might want some container images stored in Pulp. I may not want this content stored within Katello where the Katello user can mess with it (or get confused about what it is doing there). I want it stored within Pulp without Katello’s knowledge. Right now this is problematic, since Katello assumes it is the sole manager of Pulp within Foreman. For example, Katello might have some rake tasks operating on all of Pulp in order to ensure consistency with the Katello DB.

My question essentially boils down to: Is this unproblematic as soon as we are on a domain enabled Pulp version? The Idea being: My plugin could simply add it’s own separate domain to Pulp, and can then do whatever it wants within this domain. Katello would simply ignore this domain and of course my plugin would ignore the Katello domain(s). Or are there plans for Katello to use domains in a way that would prevent this?

Phrasing things differently: Are there any plans for Katello to make use of domains, and if yes, can we make it so it would still be possible to add domains that Katello “must not interact with”?

[iballou]

@quba42 thanks for bringing this up, Katello is going to have to pick up the use of domains for the future move to Pulp 4. Let me think about this a bit and I can put some opinions into my Pulpcon talk. We don’t have any plans at the moment, so now it’s a great time to collect feedback.

I like the idea of using domains to limit Katello’s functional scope, though. For example, imagine having a single Pulp server that multiple entities use (like Katello & smart proxies) at the same time. Domains could keep them separated from each other. Using Pulp behind Katello’s back is a perfect example too.