Keynote: Science in Cloud Native
summary
Cloud native adoption for science focuses on concrete wins and open problems: lazy pulling via remote snapshot in container d and a tar of tars (seekable tar / sql tar) gives near‑flat startup even for 15–20GB images whose workloads use <6% of bytes, reducing network and node storage pressure; onboarding HPC needs rootless containers and userns support (kind, minikube, k3s examples); advanced scheduling — priority queues, preemption, fair‑share, gang scheduling — is tackled by projects like volcano, admiralty, armada and virtual kubelet for multi‑cluster/resource federation.
AI-written summary · about
chapters (7)
- 0:00Intro
- 0:25Status
- 2:55Challenges
- 4:02Software Distribution
- 7:52Rootless Environments
- 9:53Advanced Scheduling
- 13:55Join this effort!
related talks
31mCloud Native Storage and Data: The CNCF Storage TAG Projects, Technology & Landscape
12mCloud Native Application-aware Data Management in multi-cloud
28mCloud Native Tech: The Building Blocks of Modern DB Software
25mCloud Native Data on K8s: The Good, Bad, and the Unknown
38mCloud Native Storage: The CNCF Storage TAG, Projects, Technology & Landscape
42mCloud Native Storage: The CNCF Storage TAG, Projects...
45mFrom Laptop to Cloud: Developing Cloud Native Applications with Containerized Databases
33mSponsored Session: Open Source Cloud Native Storage Solutions for Stateful Applications from ...