[
https://issues.apache.org/jira/browse/YUNIKORN-910?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17437729#comment-17437729
]
Weiwei Yang commented on YUNIKORN-910:
--------------------------------------
hi [~lowc1012]
According to:
https://github.com/apache/incubator-yunikorn-k8shim/blob/880669407e2db5c6c36a877756a2f529a3ee32f7/deployments/examples/predicates/pod-anti-affinity-example.yaml#L18-L20:
# this deployment ensures 2 pods cannot be co-located together on same node.
# if deploy this yaml on 1 node cluster, expecting 1 pod to be started and the
other should be pending,
# because 1 node(s) didn't match pod affinity/anti-affinity, 1 node(s) didn't
satisfy existing pods anti-affinity rules.
so on 1 node clsuter, it is expected there will be one pod pending; and 2 nodes
cluster we see expected behavior. Looks like this is working as expected.
For the e2e test failure, not sure if this is env related. Because we run the
e2e test automatically per PR pre-commit, we should be able to capture failures
early.
> Verify predicates functions for K8s 1.21
> ----------------------------------------
>
> Key: YUNIKORN-910
> URL: https://issues.apache.org/jira/browse/YUNIKORN-910
> Project: Apache YuniKorn
> Issue Type: Sub-task
> Components: shim - kubernetes
> Reporter: Chaoran Yu
> Assignee: Ryan Lo
> Priority: Major
> Fix For: 1.0.0
>
>
> _emphasized text_Similar to YUNIKORN-672
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
