Running I/O Intensive Workloads on Kubernetes
summary
Covers running high-throughput, I/O-intensive applications (network services, trading apps, databases) on Kubernetes by using Intel-added hardware acceleration drivers for multi-NIC configuration, CPU pinning and encryption offloading. Shows how to optimize a Kubernetes cluster for I/O workloads and how to use placement policies to deploy applications across multi-site Kubernetes clusters in private and public clouds.
AI-written summary · about
related talks
25mExtending Kubernetes for AI | Lessons Learned From Platform Engineering
34mUnlocking Performance: Topology-Aware CPU Scheduling W...
38mBeyond Kubernetes: Adapting To Spe...
34mWhy and how to build Kubernetes clusters with mixed CPU architectures
31mDeclarative Underlays: Scaling Purpose-Built Infrastructure(Clusters)...
32mMigrating 2 million CPU cores to Kubernetes
05mTowards a Sustainable World Realized by Kubernetes and Composable...
31mAccelerate AI adoption: automated cloud efficiency
