A reading and viewing list of how universities and research organizations actually run containers — plus the talks, papers, and primers this workshop draws on. Use it to see the patterns from 02 — Containers in Higher Ed and 04 — Cloud & Orchestration in the wild.
Some links are recordings and decks from community sessions and may require sign-in.
Containerized deployments on Amazon ECS, with SSM Parameter Store holding secrets and configuration. A namespace convention (custom apps → app name → environment) keeps IAM policies for container access simple, and each container declares the variables it needs — so the same image redeploys across environments with only its configuration changing.
- Session recording: https://drive.google.com/file/d/1IqQc6E1UqXXU13Tb8PfjapDiJPsnyjO9
A custom workspace orchestrator written in Go that spins up containers and their underlying infrastructure on demand and brokers networking/proxy traffic between cloud providers. Kubernetes-native by design (originally built for KubeCon).
- Session recording: https://drive.google.com/file/d/1mu8IMiGNaUfc4eeuwp-zhMsJjDSZtckn
YAML defines the orchestration — storage, memory, and CPU limits — and containers ship through CI/CD (Bitbucket Pipelines). The story emphasizes iterating on resource limits with rolling deployments rather than getting them right the first time.
- Session recording: https://drive.google.com/file/d/1Fl20qZs2rZTg-C4if0n_GHvdIzCLPwhP
Treats containers as the deployment model for "cookie-cutter" infrastructure-as-code, pursuing a SaaS-like operational experience instead of managing workload dependencies by hand — and highlights how cost-effective micro-workloads become in that model.
Makes the case for Singularity/Apptainer on on-prem HPC clusters: reproducibility, environment consistency across platforms, and applications that stay independent of the host OS — the reasons the Docker daemon was never the right fit on shared HPC.
- Video: https://drive.google.com/file/d/1IC3X6OE22wDT-XPWbAN_hxCxZdvWOit3
- Slides: https://docs.google.com/presentation/d/1T554vdlxutXMr4V3Ms8bLji4BLWQ4KM3zUnuskJLTwM
One container framework serves every service through environment-variable injection, with Parameter Store pushing configuration in and Route 53 handling DNS. Result: deployment time dropped from days to minutes, since a change is often just a variable update.
- Video: https://drive.google.com/file/d/16rKUrAkoVeHV-QJmUgC6HZSf40BsXDOC
- Slides: https://docs.google.com/presentation/d/1UvobZ86NTIqEYemIYBd_yTGkUby2iJOjl3s3ACfIvro/edit
A proof of concept running the Shibboleth Identity Provider on AWS ECS + Fargate: the motivation for moving identity to the cloud, a minimum-viable architecture, and where it could go next.
- Princeton — Intro to Apptainer: https://researchcomputing.princeton.edu/support/knowledge-base/apptainer
- University of Utah (CHPC) — Intro to containers (PDF): https://www.chpc.utah.edu/presentations/2026/containers26s.pdf
- Clemson RCD — Kubernetes for research computing (tutorial): https://docs.rcd.clemson.edu/training/workshop_catalog/advanced_hpc/research_computing_on_k8s/online_material_part1/
- University of Michigan — Managing research needs with Kubernetes (Bob Killen): https://www.youtube.com/watch?v=IOYgD6ppWrM
- UVA — Research application services portfolio: https://rc.virginia.edu/services/research-application-services
- Introduction to Singularity: containers for scientific & high-performance computing: https://www.youtube.com/watch?v=vEjLuX0ClN0
- Singularity: containers for science, reproducibility, and HPC (founder talk): https://www.youtube.com/watch?v=DA87Ba2dpNM
- Singularity keynote (Stanford workshop, PDF): https://www.hpcadvisorycouncil.com/events/2017/stanford-workshop/pdf/GMKurtzer_Singularity_Keynote_Tuesday_02072017.pdf
- Paper — Integration of Slurm and Kubernetes (UTK thesis): https://voljournals.utk.edu/cgi/viewcontent.cgi?article=16356&context=utk_gradthes
- University of Illinois — Hosting varied campus applications on Kubernetes: https://itproforum.illinois.edu/eventdesc/past-events/spring-2025/day-2-315-pm/how-we-use-kubernetes-to-host-varied-applications-on-campus/
- University of Bonn — Docker and Kubernetes for science and teaching: https://media.ccc.de/v/froscon2018-2264-docker_and_kubernetes_at_the_university_using_container_technology_for_science_and_teaching
- On-premise Kubernetes in software-engineering education — an experience report (ACM): https://dl.acm.org/doi/epdf/10.1145/3786580.3786953
- Migrating applications to Kubernetes (CMU SEI): https://www.sei.cmu.edu/blog/migrating-applications-to-kubernetes/
- UCSD Extended Studies — Kubernetes course: https://extendedstudies.ucsd.edu/courses/kubernetes-cse-41382
- Kubernetes design principles: understand the why — Saad Ali, Google: https://www.youtube.com/watch?v=ZuIQurh_kDk
- Kubernetes deconstructed: understanding K8s by breaking it down — Carson Anderson, DOMO: https://www.youtube.com/watch?v=90kZRyPcRZw
- Building Helm charts from the ground up — Amy Chen, Heptio: https://www.youtube.com/watch?v=vQX5nokoqrQ
- Kubernetes for network engineers (NVIDIA): https://developer.nvidia.com/blog/kubernetes-for-network-engineers/
- GitLab CI/CD automation (Docker, Kubernetes, Terraform): https://www.youtube.com/watch?v=zBrP8MzA5y0
- How to design a modern CI/CD pipeline: https://www.youtube.com/watch?v=KnSBNd3b0qI
- Using Docker to make local development a breeze: https://www.youtube.com/watch?v=zkMRWDQV4Tg
- Docker image best practices — from 1.2GB to 10MB: https://www.youtube.com/watch?v=t779DVjCKCs
- How operating systems actually work (kernel mode vs. user mode): https://www.youtube.com/watch?v=2beOYY4S0B8
- OpenAI — Scaling Kubernetes to 7,500 nodes: https://openai.com/index/scaling-kubernetes-to-7500-nodes
- OpenAI — Kubernetes outage postmortem: https://status.openai.com/incidents/01JMYB483C404VMPCW726E8MET
- Spotify — How we accidentally deleted all our Kube clusters with no user impact (David Xia): https://www.youtube.com/watch?v=ix0Tw8uinWs
- Spotify — Re-creating our entire backend without skipping a beat: https://www.youtube.com/watch?v=Iw5Tox2H87A
- Classifying architecture patterns: https://www.youtube.com/watch?v=Ic6q7S48Mho
- Thoras.ai — capacity and autoscaling: https://www.thoras.ai/