still respecting resource access and availability concerns. When Kubernetes runs on vSphere, the hypervisor platform also supports an underlying tier of high availability and automated placement options Resource Management Working Group – please join in • See Issue #49964 14 Using a NUMA aware hypervisor to solve issues now VM composition guidelines • Assuming you workload fits with the footprint compose a VM larger than the number of physical cores A NUMA aware hypervisor can have IO benefits too For the vSphere hypervisor, there are advanced vNUMA settings, they rarely need to be changed

still respecting resource access and availability concerns. When Kubernetes runs on vSphere, the hypervisor platform also supports an underlying tier of high availability and automated placement options Resource Management Working Group – please join in • See Issue #49964 14 Using a NUMA aware hypervisor to solve issues now VM composition guidelines • Assuming you workload fits with the footprint compose a VM larger than the number of physical cores A NUMA aware hypervisor can have IO benefits too For the vSphere hypervisor, there are advanced vNUMA settings, they rarely need to be changed

number of application dependencies that you need to install on the OS. Virtualization The hypervisor virtualizes the hardware limiting the number of hardware dependences that you need to install the same hardware Application Operating System Physical Infrastructure Containers VMware Hypervisor VMs Docker Containers User Cases 9 •Ready-to-go development •Self-service portal Developer

生态最为成熟，ts 主要前端 框架 Kubebuiler/react/nextjs/go-restful/ 开发CRD 最好框架，follow k8s 技术栈 计算 Containerd/cloud hypervisor/firecracker 短链路/强隔离/高安全性 存储 Openebs/sealfs 块隔离同时享有本地存储性能 网络 cilium 高性能，可计量 网关 Higress 稳定性，reload

GPU使用方式 实现 VCUDA 在vm中构建wrapper library以拦截GPU调用并将这些调用重定向 到宿主机执行 Amazon 将设备直接挂在到vm中 GPUvm 在Zen的hypervisor层实现了全虚拟化。为了隔离运行在物理 GPU上的多个VM，GPUvm将物理GPU分成几个部分，并将每个 部分分配给单个VM。 NVIDIA GRID 在硬件层面实现GPU虚拟化，每个容器可以绑定一个虚拟GPU