YoVDO

Exploration of Mixing Online Services and Offline Jobs Based on Volcano

Offered By: CNCF [Cloud Native Computing Foundation] via YouTube

Tags

Cloud Computing Courses Kubernetes Courses Cluster Management Courses Batch Processing Courses Cgroups Courses Kubelet Courses Volcano Courses

Course Description

Overview

Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!
Explore the innovative approach to resource optimization in cloud computing through this conference talk on mixing online services and offline jobs using Volcano. Discover how this technology addresses the challenge of low resource utilization and waste in clusters while maintaining service SLAs. Learn about Volcano's ability to combine online services and batch processing jobs in a single cluster, leveraging kubelet to inform schedulers about resource allocation discrepancies. Understand the concept of resource oversubscription and how it allows for the reuse of idle resources allocated to low-priority jobs. Gain insights into task isolation techniques using cgroups at the operating system level to ensure quality of service for critical online services. Delve into the practical applications and benefits of this mixed deployment strategy for improving overall cluster efficiency and resource management in cloud-native environments.

Syllabus

Exploration About Mixing Technology of Online Services and Offline Jobs Based... Leibo Wang & Lei Wu


Taught by

CNCF [Cloud Native Computing Foundation]

Related Courses

Kubernetes Housekeeping - Resource Management and Garbage Collection
Linux Foundation via YouTube
Colocating Hadoop YARN with Kubernetes to Save Costs on Big Data
CNCF [Cloud Native Computing Foundation] via YouTube
The Ins and Outs of the Cloud Provider in Kubernetes
CNCF [Cloud Native Computing Foundation] via YouTube
Kubernetes SIG Node: Introduction and Deep Dive
CNCF [Cloud Native Computing Foundation] via YouTube
Cgroupv2 - Coming Soon to Kubernetes Clusters
CNCF [Cloud Native Computing Foundation] via YouTube