elastic-gpu-scheduler
elastic-gpu-scheduler copied to clipboard
elastic-gpu-scheduler is a Kubernetes scheduler extender for GPU resources scheduling.
对于gpu-core分配策略的设计有不解之处 阅读scheduler和agent的代码后发现在agent中通过这段代码把pod获得的设备传入: ```go if len(devicesIDs) > 100 { for i := 0; i < len(devicesIDs)/100; i++ { devices = append(devices, &pluginapi.DeviceSpec{ ContainerPath: fmt.Sprintf("/host/dev/elastic-gpu-%s-%d", faker, i), HostPath: fmt.Sprintf("/dev/elastic-gpu-%s-%d", faker, i), Permissions:...
### Question if workload use Deployment or StatefulSet, annotation `elasticgpu.io/assumed: "true"` and ` elasticgpu.io/container-xxx: "0"` can add into pod yaml. But , use Pod, annotation can not add.
看着是有相关的分支的,有配置说明吗?谢谢
k8s version: 1.20.4 os: centos7 use master branch k8s have 3 worker node,only one have nvidia t4 device  schedule error log: I0222 08:58:21.218509 1 main.go:44] priority algorithm: binpack I0222...