CloudOpen at Open Source Summit Europe 2024
2024-09-16..2024-09-18
Vienna, Austria
who it's for
site reliability engineers, platform engineers, kubernetes operators, cloud architects, security engineers, observability engineers, devops engineers
topics
multi-tenant Kafka on Kubernetes; SPIFFE workload identities; Istio ambient mesh mTLS; OpenTelemetry tail sampling; RoCE/RDMA for GPU training; OVS+OVN Kubernetes CNI hardware offload; K8GB GeoDNS controller with EDNS; Unikraft unikernels scale-to-zero; secure boot TPM unified system image; OpenFGA relationship-based access control; eBPF kernel observability
| title | author | length |
|---|---|---|
| Painless Multi-Tenant Kafka on Kubernetes with Istio at | Thomas Reichel, Dominique Chanet | 47m |
ASML runs a vanilla Kafka cluster on Kubernetes using K operator and Istio with SPIFFE workload identities so clients get zero‑config authentication. A small… | ||
| Learning from Firefighters to Improve Systems Reliability | Kerim Satirli | 43m |
Adapts the Ten Standard Firefighting Orders (codified since 1957) into concrete SRE practices: keep a 120‑page pocket playbook with prerequisites, run 3–5… | ||
| Scalable Multi-Node AI Workloads in Multi-Tenant AI Clouds U... | Girish Moodalbail, Leonid Grossman | 39m |
Claim: multi-node GPU training runs at near line rate on Ethernet by combining RoCE/RDMA with OVS+OVN+OVN-Kubernetes CNI and careful topology. Implementation… | ||
| Multi-Cloud Global Content Distribution at Cloud Native Speeds | Jiri Kremser, Yury Tsarev | 40m |
K8GB claims to replace proprietary GSLB with an open source, cloud‑native GeoDNS controller that runs next to workloads (it has no master cluster), uses an… | ||
| Lightning Talk: With Great Traces Comes Great Costs: How to Reduce That Bill? | Prashansa Kulshrestha | 12m |
Distributed tracing costs explode when every span is sent to an external APM and storage/egress scale with requests; pragmatic solution uses custom sampling to… | ||
| Lightning Talk: Millisecond Scale-to-Zero and the Quest to Never Pay for Idle Again | Felipe Huici | 10m |
Claims millisecond end-to-end scale-to-zero for real apps using the Linux Foundation Unikraft project and a craft CLI that builds unikernels from Dockerfiles… | ||
| Lightning Talk: Fort Kairos: A New Dawn for Secure Linux in Untrusted Environments | Mauro Morales | 10m |
Kairos implements a trusted-boot stack for edge Linux using Secure Boot, Trusted Boot, TPM, disk encryption and boots from a unified system image signed with… | ||
| Doing for Sustainability, What Open Source Did for Software | Asim Hussain | 49m |
An open source tool named Impact Framework converts observable metrics like CPU utilization, page views, installs and prompts into emissions (carbon, water… | ||
| How to Make Your Kubernetes Add-Ons Management Painless in Multi Cloud | Eleni Grosdouli, Gianlu | 37m |
Velos runs in a management cluster and uses the Flux source controller to pull manifests, then pushes Helm charts, raw YAML/JSON and Kustomize outputs to… | ||
| Panel Discussion: eBPF: A New Era in Cloud Infrastructure Tools | Liz Rice, Frederic Branczyk | 40m |
Panelists explain what eBPF actually does (run sandboxed programs in the kernel), show real results and tradeoffs, and name tools you can use: 6 million file… | ||
| Demystifying Secure Application Communication with Zero Trust and Istio Without Sidecars | Lin Sun | 37m |
Demonstrates that Istio ambient mesh gives mutual TLS without sidecars by simply label the namespace — no pod restarts. Uses k3s local clusters with flannel… | ||
| Fine-Grained Policies RBAC with OpenFGA | José Carlos Chávez | 33m |
OpenFGA implements ReBAC (relationship‑based access control) as an open source authorization system (born at Okta, now CNCF incubation) and claims low‑latency… | ||
| Sponsored Session: How We Built GLIDE | Mickey Hoter, Madelyn Olson | 40m |
GLIDE is a multi-language Valkey client with a single Rust core and thin language wrappers that talk over a communication layer; it supports Redis OSS 6.2… | ||
| Sponsored Session: Confidential Computing - New Capabilities for New Workloads | Mike Bursell | 42m |
Hardware-based trusted execution environments run code isolated from host and other workloads and produce cryptographic measurements that an attestation… | ||
| Sponsored Session: Building the IDE Golden Path | Ben Potter | 46m |
Platform teams build a reproducible IDE golden path by shipping universal image container workspaces (backed by Dev Containers), offering a persistent home… | ||
| Cloud Native Threat Intelligence for... | Constanze Roedig, Tobias Grantner, Lukas Mahler, Josef Taha | 39m |
Kubernetes Storm Centre runs federated honey-clusters that collect eBPF and app logs, normalizes them to STIX observables, and maps matches via SQL/regex… | ||
| Panel Discussion: Do One Thing, and Do It Well: Special Purpose OSes Apply the Unix App Philosoph... | Danielle Tal, Sean McGinnis, Felipe Huici, Frederic Crozat, Erik Nordmark | 48m |
Panelists compare concrete tradeoffs: Unikraft builds modular unikernels that boot in milliseconds and run unmodified ELF binaries via a POSIX API; Project EVE… | ||
| Evolving GitOps: Harnessing Kubernetes Resource Model for 5G Core | Ashan Senevirathne, Joel Studler | 43m |
Swisscom claims that shifting GitOps to a Kubernetes Resource Model enables intent-driven 5G core configs so operators check in only high-level intent while… | ||
| Dude, Where’s My Error?: How OpenTelemetry Records Errors, and Why It... | Adriana Villela, Reese Lee | 33m |
A runnable demo shows a Python Flask app that throws an exception on even die rolls, instruments spans and span events, and sends traces, logs and metrics… | ||
| nothing matches | ||