DevConf.US 2024
| title | author | length |
|---|---|---|
| Keynote: Fireside Chat with Kelsey Hightower | Kelsey Hightower | 43m |
Career advice and concrete attitudes: start where you are, finish small projects, and build reputation by learning the job’s tech end-to-end; real claim is… | ||
| Keynote: Operations and AI with Jen Krieger | Jen Krieger | 49m |
Industry moves from six‑month button tweaks and 18‑month model updates to teams that deploy less than one day (DORA labels ~18% elite); GitHub reports… | ||
| Keynote: Student and Intern Showcase! | Urvashi Mohnani, Meera Malhotra, Austin Jamias | 54m |
Seven UMass Lowell SoarCS projects plus Red Hat intern work show concrete builds and results: QuickSwitch copies entire playlists between services via OAuth… | ||
| APIs Without Borders: Exploring the world of Locationless API Management | Vamsi Ravula | 34m |
Locationless API management uses Scupper routers and tokens to create a bidirectional Layer‑7 mesh that makes remote endpoints appear local to a single… | ||
| Auto-Instrumenting Go Libraries for Tracing with eBPF and OpenTelemetry | Mike Dame, Tyler Yahn | 39m |
An open source agent auto-instruments compiled Go binaries at runtime by analyzing the target executable, loading precompiled eBPF programs and inserting… | ||
| A Novel Approach to Generating Terraform Configurations for Google Cloud Platform | George Trammell, Max Karambelas | 38m |
Claim: a RAG system plus a custom self-querying retriever and vector DB produce a deployable single-file Terraform for Google Cloud. Implementation uses… | ||
| Everybody is talking about Confidential Computing - the minimum Devs should know | Yash Mankad, Klaus Heinrich Kiwi | 36m |
Argues that hardware-based TEEs (not FHE or SMPC) are the practical path now, using Intel TDX and AMD SEV-SNP to encrypt memory and provide remote attestation… | ||
| Expanding on a Modern Java Developer's Toolkit | Ming Wang | 24m |
Cryostat is a container-native tool that uses JDK Flight Recorder to profile Java apps running in Kubernetes and Docker, auto-discovers targets via… | ||
| Going from containers, to pods, to Kubernetes – help for your dev environments | Cedric Clyburn | 29m |
Live demo shows Podman Desktop (GUI) plus Podman CLI bridging local-to-Kubernetes workflows: builds a multi-stage Dockerfile image, runs a Python hit counter… | ||
| Handling 100000+ visitors at the world's largest physics lab (CERN) | Cristian Schuszter | 27m |
CERN builds a Science Gateway booking system that handles 100,000+ visitors by encoding visit rules declaratively and executing them with Drools so rule… | ||
| Designing Inclusive and Accessible Experiences in the Era of AI | Lisa Lyman | 35m |
Companies refuse to sell predictive policing models after 2020 concerns about bias, a 2016 chatbot (Tay) quickly turns toxic, and a Tokyo lab reports an… | ||
| Thinking like a site reliability engineer to improve CI | Brenton Leanhardt | 36m |
OpenShift CI applies SRE practices to a billion test runs per quarter, using a diff SLO and a bespoke bisect tool that diffs 100+ repos plus the OS to catch… | ||
| Jamming with CRI-O & crun: Facilitating the Convergence of AI, WASM, and Kubernetes | Peter Hunt | 31m |
CRI-O + crun run native WASM workloads in Kubernetes by detecting OCI images with platform wasi/wasm32 and dispatching them to crun with WASM plugin… | ||
| Modernization 101: A Beginner's Guide to Application Modernization and Methodology | Savitha Raghunathan, Shawn Hurley | 38m |
Claims a repeatable pipeline for large-scale app modernization using Konveyor/Conveyor (CNCF Sandbox) to analyze thousands of apps, surface incidents and… | ||
| Observability and instrumentation via OpenTelemetry | Karl Johan Grahn | 24m |
OpenTelemetry basics map logs, metrics, traces and baggage to concrete actions: a single HTTP add request generates multiple spans (route span plus a child… | ||
| Porting and Generalizing Dynamic Privilege in Linux | Arlo Albelli | 26m |
Project ports a Dynamic Privilege mechanism from x86_64 to ARM64 by adding a K_elevate syscall and a task_struct flag to track elevation, with a working ARM64… | ||
| REST & GraphQL: Adventures in schema-first API development | Jan Koscielniak, Erik Mravec | 30m |
Pyxis is a Python MongoDB service serving millions of requests per day with ~100 external and ~150 internal endpoints; it moves from duplicated OpenAPI… | ||
| RHEL 10 Roadmap | Scott McCarty | 34m |
24 months of PM interviews and customer testing produce four concrete RHEL 10 pillars: tooling, content, policy, and incremental hardening. Deliverables… | ||
| Transitioning from Debezium ElasticSearch to Postgres for Streamlined Resilience | Chris Hambridge | 33m |
Migrates a Debezium ElasticSearch pipeline into Postgres using logical replication as the chosen replacement, moving away from Kafka-backed Debezium and heavy… | ||
| UKELELE | Eric Munson | 30m |
UKELELE links application and kernel via UKL patches to register function-pointer upcalls for IO events, replacing poll/select style marshalling with direct… | ||
| UX Research and Design: Crucial to development | Amber Asaro, Vince Conzola | 33m |
Claims that great UI can boost conversion ~200% and great UX ~400% come with concrete examples: a product flub costing $70M (transparent leggings), a tie-in… | ||
| Zero-instrumentation observability based on eBPF | Peter Zaitsev | 37m |
Claims that eBPF collects metrics, logs and traces for containers and hosts with low overhead, no app code changes, and minute-scale deployment via a per-node… | ||
| ORAS: Powering the next generation of Cloud Native | Andrew Block | 49m |
ORAS provides a CLI and libraries that treat arbitrary files as OCI artifacts and actually works: demo pushes a Helm chart, updates it to v0.1.1 with oras, and… | ||
| Overcoming Challenges in Open Source Community Management | Carles Arnal | 36m |
Apicurio runs two projects: a web API designer and Apicurio Registry for storing OpenAPI, protobuf and JSON Schema artifacts; the project uses GitHub issues… | ||
| Building the Community Enterprise Operating System through CentOS Stream | Neal Gompa, Davide Cavalca | 35m |
CentOS Stream is the live stabilization pipeline feeding RHEL and currently running a public development branch CentOS Stream 10 with daily ELN builds, weekly… | ||
| Building the Responsible Workforce of Tomorrow through PIT | Colette Basiliere | 37m |
A new regional consortium runs a six-week, full-time summer fellowship that places undergraduates on six community projects and delivers maintainable… | ||
| Exploring the English Divide: A look at Open Source Inclusion | Karen Noel, Eric Duen | 36m |
A survey of 427 responses from a Red Hat platform group shows concrete gaps: 64% ESL respondents versus 36% native English, 26% live outside their birth… | ||
| Gotta go fast: how we started the Ubuntu High-Performance Computing team | Jason C. Nucciarone | 31m |
A community forms after the 2022 Ubuntu Summit to stop duplicating work and build a reusable open source HPC stack called Charmed HPC; teams combine Slurm… | ||
| How To Win Friends & Influence LLMs (with Prompt Engineering) | James Busche | 26m |
Hands-on demo of Watson X Prompt Lab shows 11 exercises and a working sandbox that connects to models including Granite model, Flan and Llama 70b; free-form… | ||
| Learn about Linux, containers, and networking through self-hosting | Justin Sun | 37m |
Chooses a small Mini PC and upgrades to Mini PC n100 with 16GB RAM and 2TB SSD, installs Fedora 40 workstation and runs containers with podman + podman-compose… | ||
| Linux Kernel Dynamic CPU Isolation | Waiman Long | 25m |
Speaker(s): Waiman Long
---
The traditional way to run a latency sensitive and CPU intensive
user space workload is to use the ""nohz_full"", ""isolcpus"" and… | ||
| Using the Open Source Mindset in the K-12 Education Space | Mary Shakshober, Kevin Hatchoua, PAULINE KIMSOUNG, Matthew Crossman, Malica Armand, Russell Lamberti, Julia Denham | 01h 10m |
Tech companies partner with Boston Public Schools through field trips, a six‑week Illustrator / UX series and a high‑school internship program that leads to… | ||
| Towards stronger checks on dependencies and compose inputs in Fedora | Adam Williamson | 29m |
Fedora is increasing gating on installability and dependency checks using openQA, Fedora CI and Testing Farm but many tests are flaky or noisy so gating isn’t… | ||
| Things I Wish I Knew When I Became a Manager | Greg Blomquist | 31m |
Distills 9+ years of engineering management into concrete tactics across Managing Teams, Managing Individuals, Managing Yourself: use a repeatable one-on-one… | ||
| To cooperate or to betray? Decoding the tester's dilemma | Deepak Koul | 30m |
Argues that the Prisoner’s Dilemma maps to five concrete tester behaviours: choosing infrastructure work (new pipeline) over actual test coverage, picking… | ||
| bpftrace: what is it, what's new, and where is it going | Jordan Rome | 28m |
A lightweight CLI language that compiles to eBPF, bpftrace runs live kernel and user‑space probes with far less code — the execsnoop example shows a one‑line… | ||
| (Less Than) 50 Ways to Build Multi-Arch Containers | Urvashi Mohnani, Adam Kaplan | 28m |
Claims there are 49 ways to build multi-arch images and demos using specific tools: Co (Go cross-compilation into a static binary on UBI micro), podman… | ||
| A Guide to Responsible Data Collection In Open Source | Arjun Devarajan | 32m |
Argues that responsible telemetry is both feasible and necessary for open source, demonstrates work with Scarf to collect download and runtime usage (npm… | ||
| How Generative AI and Automation Can Revolutionize Certification Study | Randy Romero, Angela Andrews, Jordan Jacobs, Kush Gupta | 33m |
Shows how to turn certification prep into repeatable, testable work: build ephemeral labs with Proxmox, Terraform or Ansible and spin identical VMs in AWS; use… | ||
| AI Lab Recipes: Cooking up AI Applications on your Laptop | Michael Clifford | 37m |
Open-source GitHub repo ai-lab-recipes packages containerized AI apps so a developer can spin up a local code generation assistant in two commands using the… | ||
| Building Trust with LLMs | Surya Pathak, Hema Veeradhi | 34m |
Compares three concrete evaluation paths: automatic metrics like ROUGE and BLEU and F1 which miss semantics, model‑based scoring using LangChain criteria… | ||
| Creating Your Own LLM Tuning Platform with Open Source Technologies | James Busche, Kelly Abuelsaad | 35m |
An open source pipeline called FMS HF Tuning wraps HuggingFace SFT Trainer and PyTorch/FSDE for distributed runs, exposes a CLI sft_trainer.py and a Jupyter… | ||
| Data Security and Storage Hardening in Rook and Ceph | Federico Lucifredi, Sage McTaggart | 35m |
Rook deploys Ceph on Kubernetes and claims hardening becomes a YAML switch; the talk shows concrete options: 6.5 exabytes of deployed Ceph, FIPS 140-2 cipher… | ||
| Efficiently Deploying and Benchmarking LLMs in Kubernetes | Nikhil Palaskar | 30m |
Deploys LLMs on OpenShift AI using the KServe stack and automates installs, CRDs and CI with topsell; a multiprocessing Python tool llm-load-test simulates… | ||
| Jumpstarter: Cloud-Native Hardware in the Loop | Nick Cao | 24m |
Jumpstarter is a Python framework plus an exporter daemon and cloud controller that turns physical boards into leaseable test resources, exposing power, GPIO… | ||
| Jupyter extension for executing kubeflow pipeline seamlessly | Harshad Reddy Nalla | 27m |
Elyra ties Jupyter to Kubeflow Pipelines v2 so notebooks become pipeline components via a visual drag‑and‑drop editor and one‑click pipeline submission. The… | ||
| Open Education in the New England Research Cloud (NERC) | Isaiah Stapleton | 27m |
Open Education Project builds Jupyter Book textbooks and containerized JupyterLab environments, runs them on the New England Research Cloud using OpenShift AI… | ||
| Optimizing your Hybrid Cloud Operating System Experience | Yu Qi (Jerry) Zhang, Ines Qian | 32m |
Describes a move from Ansible/RPM node-by-node upgrades to an operator-driven flow that builds and pushes per-cluster OS artifacts via an on-cluster image… | ||
| Truth-seeker: Using LLM agents to build and verify knowledge bases | Jeremy Peterson | 32m |
Open-source LLM agents extract statements from documents, run web searches to score statements for sourcing and consistency, and tag each node as fact… | ||
| Ensuring a Consistent Installation Time for ROSA with Hosted Control Planes | Russ Zaleski | 30m |
Walks through creating thousands of hosted control planes using hcp-burner and workload driver kuerer to measure where installs stall, finds a 10→5 minute… | ||
| 10 Cool features in Podman and Podman Desktop | Daniel Walsh | 36m |
Podman 5.0 and Podman Desktop add concrete tooling: podman farm orchestrates multi‑arch builds by SSHing to remote podman services, building natively on each… | ||
| Detoxification of LLMs using TrustyAI Detoxify and HuggingFace SFTTrainer | Christina Xu | 11m |
Claims a no-annotator detox pipeline: TrustyAI Detoxify rephrase via token disagreement by scoring next-token probability differences between two T‑models… | ||
| GPU Accelerated Containers on Apple Silicon with libkrun and podman machine | Tyler Fanelli, Jake Correnti | 16m |
Shows libkrun plugged into podman machine (via krunkit) to enable Apple GPU passthrough so a local 7B inference run (mistl/llama.cpp-style model on macOS… | ||
| Nmstate Polyglot Model - Translating Natural Language into Nmstate States | Wen Liang | 38m |
The team builds an Nmstate Polyglot model that translates natural language into Nmstate YAML, fine-tunes models with LoRA fine-tuning on normalized… | ||
| Podman 5 & Long-Term Software Maintenance | Matthew Heon | 37m |
Podman 5 ships as a major, breaking release that prioritizes long-term maintainability: a 70% rewrite of podman machine raises code sharing from ~30% to… | ||
| Self-Hosted LLMs: A Practical Guide | Hema Veeradhi, Aakanksha Duggal | 35m |
Guide shows a full, working local setup: download a quantized Whisper model (~466 MB from Hugging Face), build a C++ whisper.cpp binary into a container with… | ||
| Store AI/ML models efficiently with OCI Artifacts | Tom Coufal | 35m |
Claim: OCI Artifacts store, version and distribute AI models as content-addressed blobs so updates send only differences. Uses the ORAS CLI and the mldm Python… | ||
| Super Accessible No Math Intro to Neural Networks For Beginners | Lance Galletti | 24m |
Claims that neural networks are simple math processes using only addition/subtraction; they learn by warping input data toward a fixed guess via iterative… | ||
| Who Watches the Watchmen? Understanding LLM Benchmark Quality | Erik Erlandson | 30m |
Shows that popular benchmarks like MMLU with 115,000 questions miss almost all real input space once you calculate combinatorics from a 100,000‑word vocabulary… | ||
| Your own personal supercomputer within 15 minutes or less | Jason C. Nucciarone | 32m |
Claims a full open‑source HPC stack called Charmed HPC can deploy in 15 minutes using Juju to orchestrate Slurm for scheduling, CephFS for distributed storage… | ||
| How to Build Collaboration and Influence Open Source Projects | Michael McCune | 34m |
Lessons from decade-plus open source work and observable outcomes: built a cross‑compile toolchain then moved it toward open source (boss declines), helped… | ||
| Unlock Your Team's Superpower - It's Not in Your Toolbox, It's in Your Head! | Nancy Jain, Anuj Singla | 36m |
Covers Google’s Project Aristotle claim that team dynamics beat individual hires, citing analysis of ~180 teams and ~250 attributes and extracting five… | ||
| Accelerating Linux Boot Time: Techniques and Strategies for Optimal Performance | Eric Curtin, Ed Chong, Brian Masney | 37m |
Measurements combine systemd-analyze and CNTVCT counters plus dmesg/journal parsing to separate firmware, kernel pre-timer and userspace; pre-timer falls from… | ||
| Automate Openshift Cluster deployment with RHACM and AAP | Michael DiDato, Michael Zamot, Michael Navarro | 31m |
Deploy OpenShift clusters by having ACM generate managed-cluster objects and ISOs from GitLab-hosted Ansible templates, then trigger AAP via webhook and a… | ||
| Building a Better Software Supply Chain | Ann Marie Fred | 27m |
A monolithic internal build system with 250 services across 1000 hosts is replaced by an open source Konflux platform on Kubernetes using Argo CD, Tekton with… | ||
| Connecting the Dots: Skupper.io as an application modernization Enabler | Vamsi Ravula | 37m |
Skupper.io builds a layer-7 overlay that lets services keep stable names instead of IPs and uses routers plus tokens to form a mesh; setup uses the CLI… | ||
| Container life cycle management in Automotive/Edge deployments. | Yariv Rachmani, Rakesh Musalay | 37m |
Explains how Podman’s Quadlet replaces manual unit crafting by turning simple declarative .container/.network/.volume qulet files into systemd units… | ||
| Crafting Seamless Development: Improving Kubernetes Operator Developer Experience | Zack Zlotnik, Yu Qi (Jerry) Zhang | 27m |
Work on the Machine Config Operator for OpenShift and Go cuts a slow “edit-compile-run-debug” cycle into concrete fixes: initial sandbox clusters take 30–60… | ||
| DOWN with Shift Left: Why You Should be Shifting DOWN into the Platform | Scott Rosenberg | 28m |
Argues that shift left simply shifts cognitive load onto developers and instead promotes shift down to the platform, where automation and policy enforce… | ||
| Paving the Way for Uninterrupted Car Operations | Yariv Rachmani, Douglas Schilling Landgraf | 36m |
FFI on containers with podman separates ASIL-critical services into an uninterfered partition and noncritical workload into a QM partition using namespaces… | ||
| Improve Upstream Code Quality by Bringing Testing to Patch Level | Lei Yang, Yumei Huang | 26m |
Downstream testers move regression testing upstream by subscribing to mailing lists, cloning upstream repos, downloading patch series with b4m, applying with… | ||
| KubeArchive: Don’t use Kubernetes as a Database | Sam Koved, Greg Allen | 35m |
KubeArchive streams Kubernetes objects to a real database (demo uses Postgres) in real time via KNative Eventing API ServerSource, evaluates user-provided CEL… | ||
| Automating lifecycle of virtual machines in Kubernetes using Ansible and KubeVirt | Andrew Block, Harsha Cherukuri | 32m |
Claim: Ansible automates full VM lifecycle inside Kubernetes using Ansible Collections and KubeVirt/OpenShift Virtualization primitives. Demo shows an… | ||
| Linux in Cars! CentOS AutoSD, SDVs, and more | Jeffrey ‘Jefro’ Osier-Mixon | 28m |
CentOS Automotive SIG runs an AutoSD preview distribution (not certified for in‑car use) that shapes the Red Hat in‑vehicle OS roadmap, with a realtime‑patched… | ||
| Managing thousands of DNS records in a GitOps fashion using Ansible and NS1 | Michael Kearey | 31m |
Red Hat moves DNS from legacy BIND into a GitOps pipeline using GitLab, Ansible Automation Platform, and NS1: an ns1 Ansible role plus a set of playbooks and a… | ||
| Policy-Driven Supply Chain Security with Enterprise Contract | Mark Bestavros | 36m |
Enterprise Contract runs a policy engine that verifies container images’ signatures and attestations and gates releases; the demo uses cosign (both key and… | ||
| Power Efficiency Aware Kubernetes Scheduler | Han Dong | 29m |
PEAKS is a Kubernetes scheduler that uses Kepler (eBPF → Prometheus) metrics and an online multi‑objective learner (A.D) to pick pod replica counts and node… | ||
| SBOMs: zero to hero in 25 mins! | Brian Cook | 34m |
Crash course gives concrete workflow: use cachy2 to prefetch and lock RPMs, generate an SBOM from that cache (UBI9 build: 192 RPMs, sift reports 210), run sift… | ||
| Scale your Batch / Big Data / AI Workloads Beyond the Kubernetes Scheduler | Anish Asthana, Kevin Postlethwait | 25m |
Benchmarks run on a 100-node simulated cluster show the default kube-scheduler fails to finish 1,000 jobs × 10 pods (only ~51% complete, massive rescheduling… | ||
| Serverless Java in Action: Cloud Agnostic Design Patterns and Tips | Daniel Oh | 31m |
Claims a cloud‑agnostic, production serverless Java stack built with Quarkus + Knative, using native compilation to cut startup to milliseconds. Demo wires a… | ||
| SpiceDB: open source, hyperscale authorization | Jimmy Zelinskie | 35m |
Open source SpiceDB implements a Zanzibar-inspired model to centralize permissions, run relationship traversals (ReBAC) at low latency, and answer queries like… | ||
| The OSS IaC Tooling Face-off for Modern Cloud Native Ops | Asaf Blubshtein | 24m |
Compares concrete tradeoffs between tools: Terraform dependency graph and provider ecosystem vs CloudFormation’s AWS-only, no-state model; Pulumi runs in… | ||
| Why Open Source and Web3 need to find (back) together | Daniel Riek | 38m |
Claim: Open Source has shifted toward corporate interests and cloudification creates central choke points that erode user sovereignty. Evidence: GitHub ~90%… | ||
| Working with filesystem in Time Series database | Roman Khavronenko | 35m |
VictoriaMetrics is written in Go and targets heavy ingestion, with 2–4 million samples/sec typical and single clusters exceeding 100M. Durability design trades… | ||
| What is all about? | Mike Dame | 04m |
Conference kickoff montage that insists the event centers on openness and AI plus open source, noting over 100 sessions and hundreds of contributors. Mentions… | ||
| Cloud-Native Databases in Kubernetes with OpenShift and Postgres | Torsten Steinbach, Michael St-Jean, Gabriele Bartolini | 41m |
Shows how OpenShift plus PostgreSQL (via the CloudNativePG operator) runs a three‑node cluster with one synchronous replica, leverages availability zones for… | ||
| Delegation is a Love Language | Jeff Ligon | 34m |
Argues that delegation is a measurable career lever—a trust-based “love language” that multiplies impact when handled with one-on-ones, clear asks, and… | ||
| nothing matches | ||