manager, Reduce toleration in case of node reboot #374
Conversation
|
Since kubevirtci does not support multiple master clusters, there is no option to check this behavior on CI, so I checked this PR on U/S cluster manually by
|
|
/retest |
|
/cc @ormergi |
There was a problem hiding this comment.
Hi @RamLavi, please consider adding a reference to the relevant K8s documentation to the PR description:
https://kubernetes.io/docs/concepts/scheduling-eviction/taint-and-toleration/#taint-based-evictions
Currently if the node upon which KMP-manager deployment runs reboots, the KMP pod is evicted only 5 minutes (default) after the node is unreachable. Since KMP-manager deployment is is essential for VM create/update service-ability, reducing this eviction to 1 minute. This is done by altering the tolerations config [0] This means that if the node is not available/ready for more than 1 minute, the pod will evict and will try to reschedule using the configured affinity. [0] https://kubernetes.io/docs/concepts/scheduling-eviction/taint-and-toleration/#taint-based-evictions Signed-off-by: Ram Lavi <ralavi@redhat.com>
RamLavi
force-pushed
the
reduce_kmp_node_eviction_time
branch
from
4de1d87
to
85cf418
Compare
Done |
|
@orelmisan: changing LGTM is restricted to collaborators In response to this: Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
|
failed test errored on BadRequest from the server, and it doesn't seem related. I want to rerun the lane a few times to be sure |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: qinqon The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
What this PR does / why we need it:
Currently if the node upon which KMP-manager deployment runs reboots,
the KMP pod is evicted only 5 minutes (default) after the node is unreachable.
Since KMP-manager deployment is is essential for VM create/update
service-ability, reducing this eviction to 1 minute. This is done by
altering the tolerations config [0]
This means that if the node is not available/ready for more than 1
minute, the pod will evict and will try to reschedule using the
configured affinity.
[0] https://kubernetes.io/docs/concepts/scheduling-eviction/taint-and-toleration/#taint-based-evictions
Special notes for your reviewer:
Release note: