[release-1.2] Disable VMI migration upon K8s unschedulable taint (#823) #904
Conversation
Currently, the default behaviour is to migrate the VMI when a node is tainted with the node.kubernetes.io/unschedulable key. This behaviour is wrong and can be un-expected from the user's perspective because the unschedulable taint means "do not schedule new workloads on this node" and not "evict existing workloads from this node". kubevirt/kubevirt#4012 adds support for the eviction API so we can properly handle node drains and specific evictions on VMI pods. This patch avoid setting (and explicitly removes during upgrades if set in the past) the default node drain key so we won't migrate VMIs when we're not expected to do so. Fixes: https://bugzilla.redhat.com/1881676 Signed-off-by: Daniel Belenky <dbelenky@redhat.com> Signed-off-by: Simone Tiraboschi <stirabos@redhat.com>
kubevirt-bot
added
release-note
|
All tests passed |
|
@tiraboschi: The following test failed, say
Full PR test history. Your PR dashboard. 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. I understand the commands that are listed here. |
|
/approve |
kubevirt-bot
added
the
approved
|
hco-e2e-upgrade-prev-aws lane succeeded. |
|
@hco-bot: Overrode contexts on behalf of hco-bot: ci/prow/hco-e2e-upgrade-prev-azure 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. |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: nunnatsa, tiraboschi 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 |
This is a cherry-pick of #823
Currently, the default behaviour is to migrate the VMI when a node is
tainted with the node.kubernetes.io/unschedulable key. This behaviour is
wrong and can be un-expected from the user's perspective because the
unschedulable taint means "do not schedule new workloads on this node"
and not "evict existing workloads from this node".
kubevirt/kubevirt#4012 adds support for the
eviction API so we can properly handle node drains and specific
evictions on VMI pods.
This patch avoid setting (and explicitly removes during upgrades if
set in the past) the default node drain key so we won't migrate VMIs
when we're not expected to do so.
Fixes: https://bugzilla.redhat.com/1881676
Signed-off-by: Daniel Belenky dbelenky@redhat.com
Signed-off-by: Simone Tiraboschi stirabos@redhat.com
Release note: