亲和性调度对接说明

功能介绍

通过在K8s资源的Pod模板中添加特定Label或Annotation，可控制Volcano的核心调度行为，包括但不限于：

• 昇腾AI处理器的亲和性调度
• 交换机亲和性调度
• 逻辑超节点亲和性调度
• 故障重调度

前提条件

确保Kubernetes集群已经正确部署并配置了Volcano调度器，并且相关的调度插件Ascend-for-volcano已启用。

调度策略配置示例

以StatefulSet为例，所有调度相关的labels/annotations均需配置在StatefulSet.spec.template.metadata下，确保调度器可以从Pod实例中正确读取。

apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: mindx-dls-test               # The value of this parameter must be consistent with the name of ConfigMap.
  labels:
    app: mindspore
    ring-controller.atlas: ascend-910
spec:
  replicas: 16                        # The value of replicas is 1 in a single-node scenario and N in an N-node scenario. The number of NPUs in the requests field is 8 in an N-node scenario.
  podManagementPolicy: Parallel   # 支持OrderdReady和Parallel两种模式。“OrderdReady”仅支持节点内亲和调度并且huawei.com/schedule_minAvailable只能为1。“Parallel”支持节点内和节点间亲和调度
  serviceName: service-headliness
  selector:
    matchLabels:
      app: mindspore
  template:
    metadata:
      labels:
         app: mindspore
         ring-controller.atlas: ascend-910
        fault-scheduling: force   # 故障重调度功能开关
        pod-rescheduling: "on"   # Pod级别重调度功能开关
        fault-retry-times: "85"    # 业务面故障重调度次数
        tor-affinity: large-model-schema  # 交换机亲和性调度开关
         deploy-name: mindx-dls-test # 生成rankTable必须增加该标签，取值和任务名称保持一致
      annotations:
        sp-block: "128"         # 逻辑超节点亲和性调度开关
        huawei.com/recover_policy_path: pod    # Pod级别重调度不升级为Job级开关
        huawei.com/schedule_minAvailable: "16"  # 任务调度的最小副本数，建议与任务副本数保持一致
    spec:
      schedulerName: volcano         # Use the Volcano scheduler to schedule jobs.
      nodeSelector:
        host-arch: huawei-arm        # Configure the label based on the actual job.
      containers:
        - image: ubuntu:18.04      # Training framework image, which can be modified.
          name: mindspore
          resources:
            requests:
              huawei.com/Ascend910: 16                                               # Number of required NPUs. The maximum value is 16. You can add lines below to configure resources such as memory and CPU
            limits:
              huawei.com/Ascend910: 16                                                # The value must be consistent with that in requests.

• 如果一个PodGroup被创建，则spec中的调度配置将覆盖其生成的Pod上的labels/annotations配置。
• 对于可以生成PodGroup的资源，在PodGroup上配置对应的调度策略也可以实现亲和性调度能力。
• 常用的Label与Annotation对照表请参见PodGroup/Pod。