[
https://issues.apache.org/jira/browse/FLINK-31047?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Tamir Sagi updated FLINK-31047:
-------------------------------
Description:
In case HA is enabled, it should be possible to define topology spread
constrains to make sure the replicas are deployed on different nodes/zones
Add the following to flink-operator.yaml template under spec clause
[https://github.com/apache/flink-kubernetes-operator/blob/main/helm/flink-kubernetes-operator/templates/flink-operator.yaml#L64]
topologySpreadConstraints:
- maxSkew: 1
topologyKey: topology.kubernetes.io/zone
whenUnsatisfiable: DoNotSchedule
we can also add label selector to it.
Reference:
[https://kubernetes.io/docs/concepts/scheduling-eviction/topology-spread-constraints/]
was:
In case HA is enabled, it should be possible to define topology spread
constrains to make sure the replicas are deployed on different nodes/zones
Add the following to flink-operator.yaml template under containers clause
[https://github.com/apache/flink-kubernetes-operator/blob/main/helm/flink-kubernetes-operator/templates/flink-operator.yaml#L64]
topologySpreadConstraints:
- maxSkew: 1
topologyKey: topology.kubernetes.io/zone
whenUnsatisfiable: DoNotSchedule
we can also add label selector to it.
Reference:
[https://kubernetes.io/docs/concepts/scheduling-eviction/topology-spread-constraints/]
> Support topology spread constrains when HA is enabled
> -----------------------------------------------------
>
> Key: FLINK-31047
> URL: https://issues.apache.org/jira/browse/FLINK-31047
> Project: Flink
> Issue Type: Improvement
> Components: Kubernetes Operator
> Reporter: Tamir Sagi
> Priority: Major
> Labels: flink-kubernetes-operator
>
> In case HA is enabled, it should be possible to define topology spread
> constrains to make sure the replicas are deployed on different nodes/zones
>
> Add the following to flink-operator.yaml template under spec clause
> [https://github.com/apache/flink-kubernetes-operator/blob/main/helm/flink-kubernetes-operator/templates/flink-operator.yaml#L64]
> topologySpreadConstraints:
> - maxSkew: 1
> topologyKey: topology.kubernetes.io/zone
> whenUnsatisfiable: DoNotSchedule
> we can also add label selector to it.
> Reference:
> [https://kubernetes.io/docs/concepts/scheduling-eviction/topology-spread-constraints/]
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
