Data mover pod schedule mechanism support
Specify the following fields in the backup server ConfigMap to schedule data mover pods on the nodes.
nodeSelector: nodeSelector is the effortless way to constrain pods to the nodes with specific labels.
Example:
apiVersion: v1 kind: ConfigMap metadata: name: backupserver.sample.domain.com namespace: netbackup data: datamover.hostaliases: | 10.20.12.13=backupserver.sample.domain.com 10.21.12.13=mediaserver.sample.domain.com datamover.properties: | image=reg.domain.com/datamover/image:latest datamover.nodeSelector: | kubernetes.io/hostname: test1-l94jm-worker-k49vj topology.rook.io/rack: rack1 version: "1"
nodeName: nodeName is a direct form of node selection than affinity or nodeSelector. It allows you to specify a node on which a pod is scheduled for backup, overriding the default schedule mechanism.
Example:
apiVersion: v1 kind: ConfigMap metadata: name: backupserver.sample.domain.com namespace: netbackup data: datamover.hostaliases: | 10.20.12.13=backupserver.sample.domain.com 10.21.12.13=mediaserver.sample.domain.com datamover.properties: | image=reg.domain.com/datamover/image:latest datamover.nodeName : test1-l94jm-worker-hbblk version: "1"
Taint and Toleration: Toleration allows you to schedule the pods with similar taints. Taint and toleration work together to ensure that the pods are scheduled onto appropriate nodes. If one or more taints are applied to a node. Then that node must not accept any pods which does not tolerate the taints.
Example:
apiVersion: v1 kind: ConfigMap metadata: name: backupserver.sample.domain.com namespace: netbackup data: datamover.hostaliases: | 10.20.12.13=backupserver.sample.domain.com 10.21.12.13=mediaserver.sample.domain.com datamover.properties: | image=reg.domain.com/datamover/image:latest datamover.tolerations: | - key: "dedicated" operator: "Equal" value: "experimental" effect: "NoSchedule" version: "1"
Affinity and Anti-affinity: Node affinity functions like the nodeSelector field but it is more expressive and allows you to specify soft rules. Inter-pod affinity/anti-affinity allows you to constrain pods against labels on the other pods.
Examples:
Node Affinity:
apiVersion: v1 kind: ConfigMap metadata: name: backupserver.sample.domain.com namespace: netbackup data: datamover.hostaliases: | 10.20.12.13=backupserver.sample.domain.com 10.21.12.13=mediaserver.sample.domain.com datamover.properties: | image=reg.domain.com/datamover/image:latest datamover.affinity: | nodeAffinity: requiredDuringSchedulingIgnoredDuringExecution: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - test1-l94jm-worker-hbblk preferredDuringSchedulingIgnoredDuringExecution: - weight: 1 preference: matchExpressions: - key: beta.kubernetes.io/arch operator: In values: - amd64 version: "1"
Pod Affinity
apiVersion: v1 kind: ConfigMap metadata: name: backupserver.sample.domain.com namespace: netbackup data: datamover.hostaliases: | 10.20.12.13=backupserver.sample.domain.com 10.21.12.13=mediaserver.sample.domain.com datamover.properties: | image=reg.domain.com/datamover/image:latest datamover.affinity: | podAffinity: requiredDuringSchedulingIgnoredDuringExecution: - labelSelector: matchExpressions: - key: component operator: In values: - netbackup topologyKey: kubernetes.io/hostname version: "1"
topologySpreadConstraints: Topology spread constraints are used to control the behavior of the pods that are spread across your cluster among failure-domains such as regions, zones, nodes, and other user-defined topology domains.
Example:
apiVersion: v1 kind: ConfigMap metadata: name: backupserver.sample.domain.com namespace: netbackup data: datamover. hostaliases: | 10.20.12.13=backupserver.sample.domain.com 10.21.12.13=mediaserver.sample.domain.com datamover.properties: | image=reg.domain.com/datamover/image:latest datamover.topologySpreadConstraints : | - maxSkew: 1 topologyKey: kubernetes.io/hostname whenUnsatisfiable: DoNotSchedule version: "1"
Labels: Labels are the key/value pairs attached to the objects, such as pods. Labels intends to identify the attributes of an object which are significant and relevant to users. Labels can organize and select subsets of objects. Labels which are attached to objects at creation time are subsequently added and modified at any time.
Example:
apiVersion: v1 kind: ConfigMap metadata: name: backupserver.sample.domain.com namespace: netbackup data: datamover.hostaliases: | 10.20.12.13=backupserver.sample.domain.com 10.21.12.13=mediaserver.sample.domain.com datamover.properties: | image=reg.domain.com/datamover/image:latest datamover.labels: | env: test pod: datamover version: "1"
Annotations: User can use either labels or annotations to attach metadata to Kubernetes objects. You cannot use Annotations to identify and select objects.
Example:
apiVersion: v1 kind: ConfigMap metadata: name: backupserver.sample.domain.com namespace: netbackup data: datamover.hostaliases: | 10.20.12.13=backupserver.sample.domain.com 10.21.12.13=mediaserver.sample.domain.com datamover.properties: | image=reg.domain.com/datamover/image:latest datamover.annotations: | buildinfo: |- [{ "name": "test", "build": "1" }] imageregistry: "https://reg.domain.com/" version: "1"