How Container Schedulers and Software-based Storage will Change the Cloud
summary
Claim: a Mesos ScaleIO framework wires software-defined storage, REX-Ray and the Mesos module into every agent so persistent containers get high availability and automated day‑2 ops. Demo deploys ScaleIO on three agents in AWS, fills storage to 98% full, then framework calls AWS APIs to create and attach three 1TB EBS volumes and tells ScaleIO to rebalance, dropping used capacity to ~14%. The offer‑accept scheduling model plus storage APIs actually provisions, attaches and heals storage automatically.
AI-written summary · about
chapters (19)
- 0:00Start
- 1:16What are they?
- 1:58What makes them unique?
- 4:03Example: NFS
- 4:32Example: VSAN
- 6:05What is a Scheduler?
- 7:34Scheduling Work
- 9:16Custom Scheduling
- 10:24Apache Mesos
- 11:21Mesos Frameworks
- 13:57Framework/Offer Mechanism
- 19:34SDS Framework = Mind Blow
- 22:34Containers Today
- 24:57Death of a Container
- 27:54Manages the Storage Enablement
- 29:23What this Means for your Apps
- 29:51Moving towards the Cloud
- 30:52Self-aware Applications
- 33:15The Moving Parts
related talks
44mImproving Storage Support in Apache Mesos using CSI
39mHow we learned to love the Data Center Operating System
25mContainer Attached Storage: Open Source Extensions: Kubernetes + Data
53mDocker as the Building Block for Datacenter-Scale Applications
42mMesos: The Operating System for your Cluster
08mGold Sponsor Lightning Talk: The Future of Container Storage: CSI and Cloud Foundry
01h 04mLaunching patterns for containers - it's more than just scheduling
39mFill a Gap of Kubernetes: IO Resource Scheduling and Isolation