Add force Daemon Sets option #5444
Conversation
k8s-ci-robot
added
the
kind/feature
k8s-ci-robot
added
area/cluster-autoscaler
cncf-cla: yes
BigDarkClown
force-pushed
the
force-schedule-ds
branch
from
11c7e53 to
908556e
Compare
k8s-ci-robot
removed
the
needs-rebase
BigDarkClown
force-pushed
the
force-schedule-ds
branch
from
908556e to
903c619
Compare
|
/assign |
| if pod_util.IsMirrorPod(pod) || pod_util.IsDaemonSetPod(pod) { | ||
| selectedPods = append(selectedPods, allPods[id]) | ||
| // Add scheduled mirror pods and scheduled DS pods if not force scheduling DS | ||
| if pod_util.IsMirrorPod(pod) || (!forceDaemonSets && pod_util.IsDaemonSetPod(pod)) { |
There was a problem hiding this comment.
This will ignore all DaemonSet-like (but not really DaemonSet) pods for which IsDaemonSetPod currently returns true. I think we need to distinguish between "real DS" and "DS-like" pods here to make this function work correctly.
There was a problem hiding this comment.
That is a good point. I think we can remedy this by checking which DS pods are already present on the chosen node and inserting new DS pods only for DSs that are not present. That way we will still keep all the "DS-like" pods to the template.
BigDarkClown
force-pushed
the
force-schedule-ds
branch
from
903c619 to
b608278
Compare
|
/hold Need to test this change some more |
k8s-ci-robot
added
the
do-not-merge/hold
|
/unhold |
k8s-ci-robot
removed
the
do-not-merge/hold
|
/lgtm |
k8s-ci-robot
added
the
lgtm
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: BigDarkClown, x13n 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 |
k8s-ci-robot
added
the
approved
What type of PR is this?
/kind feature
What this PR does / why we need it:
This PR adds an optional feature that force-schedules all Daemon Set pods on node group templates. That means that node groups which cannot schedule all the expected daemon sets (eg. because of small nodes) will no longer be chosen during scale-up.
The change is disabled and guarded by a manifest flag, so it should not have any effect on the component yet.