Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

update code docs around old todo that is not going to happen #79892

Merged
merged 1 commit into from
Jul 11, 2019
Merged

update code docs around old todo that is not going to happen #79892

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 5 additions & 7 deletions pkg/kubelet/status/status_manager.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -582,13 +582,11 @@ func (m *manager) needsReconcile(uid types.UID, status v1.PodStatus) bool {
return true
}

// We add this function, because apiserver only supports *RFC3339* now, which means that the timestamp returned by
// apiserver has no nanosecond information. However, the timestamp returned by metav1.Now() contains nanosecond,
// so when we do comparison between status from apiserver and cached status, isPodStatusByKubeletEqual() will always return false.
// There is related issue #15262 and PR #15263 about this.
// In fact, the best way to solve this is to do it on api side. However, for now, we normalize the status locally in
// kubelet temporarily.
// TODO(random-liu): Remove timestamp related logic after apiserver supports nanosecond or makes it consistent.
// normalizeStatus normalizes nanosecond precision timestamps in podStatus
// down to second precision (*RFC339NANO* -> *RFC3339*). This must be done
// before comparing podStatus to the status returned by apiserver because
// apiserver does not support RFC339NANO.
// Related issue #15262/PR #15263 to move apiserver to RFC339NANO is closed.
func normalizeStatus(pod *v1.Pod, status *v1.PodStatus) *v1.PodStatus {
bytesPerStatus := kubecontainer.MaxPodTerminationMessageLogLength
if containers := len(pod.Spec.Containers) + len(pod.Spec.InitContainers); containers > 0 {
Expand Down
2 changes: 1 addition & 1 deletion pkg/kubelet/status/status_manager_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -745,7 +745,7 @@ func TestReconcilePodStatus(t *testing.T) {
// If the pod status is the same, only the timestamp is in Rfc3339 format (lower precision without nanosecond),
// a reconciliation is not needed, syncBatch should do nothing.
// The StartTime should have been set in SetPodStatus().
// TODO(random-liu): Remove this later when api becomes consistent for timestamp.
// This test is done because the related issue #15262/PR #15263 to move apiserver to RFC339NANO is closed.
t.Logf("Syncbatch should do nothing, as a reconciliation is not required")
normalizedStartTime := testPod.Status.StartTime.Rfc3339Copy()
testPod.Status.StartTime = &normalizedStartTime
Expand Down