Skip to content

OCPNODE-3659: Not fail upgrade checks if all nodes are ready #30318

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

Merged
merged 1 commit into from
Oct 2, 2025
Merged

OCPNODE-3659: Not fail upgrade checks if all nodes are ready #30318

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
18 changes: 17 additions & 1 deletion test/e2e/upgrade/upgrade.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -41,6 +41,7 @@ import (
"github.com/openshift/origin/test/e2e/upgrade/adminack"
"github.com/openshift/origin/test/e2e/upgrade/dns"
"github.com/openshift/origin/test/e2e/upgrade/manifestdelete"
mc "github.com/openshift/origin/test/extended/machine_config"
"github.com/openshift/origin/test/extended/prometheus"
"github.com/openshift/origin/test/extended/util/disruption"
"github.com/openshift/origin/test/extended/util/operator"
Expand Down Expand Up @@ -625,6 +626,21 @@ func clusterUpgrade(f *framework.Framework, c configv1client.Interface, dc dynam
func() (error, bool) {
framework.Logf("Waiting on pools to be upgraded")
if err := wait.PollImmediate(10*time.Second, 30*time.Minute, func() (bool, error) {

nodes, err := kubeClient.CoreV1().Nodes().List(context.Background(), metav1.ListOptions{})
Copy link
Member

@wking wking Oct 1, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's hard to imagine a MachineConfigPool update that cordons and drains control-plane nodes where we never see a single master Node that's Unschedulable=True in this 10s poll loop, so looks good to me.

if err != nil {
framework.Logf("error getting nodes %v", err)
return false, nil
}

allNodesReady := true
for _, node := range nodes.Items {
if !mc.IsNodeReady(node) {
allNodesReady = false
break
}
}

mcps := dc.Resource(schema.GroupVersionResource{
Group: "machineconfiguration.openshift.io",
Version: "v1",
Expand All @@ -641,7 +657,7 @@ func clusterUpgrade(f *framework.Framework, c configv1client.Interface, dc dynam
allUpdated = allUpdated && updated

// Invariant: when CVO reaches level, MCO is required to have rolled out control plane updates
if p.GetName() == "master" && requiresUpdate && errMasterUpdating == nil {
if p.GetName() == "master" && requiresUpdate && !allNodesReady && errMasterUpdating == nil {
errMasterUpdating = fmt.Errorf("the %q pool should be updated before the CVO reports available at the new version", p.GetName())
framework.Logf("Invariant violation detected: %s", errMasterUpdating)
}
Expand Down
8 changes: 4 additions & 4 deletions test/extended/machine_config/helpers.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -494,7 +494,7 @@ func ValidateMCNForNodeInPool(oc *exutil.CLI, clientSet *machineconfigclient.Cli
// `GetRandomNode` gets a random node from with a given role and checks whether the node is ready. If no
// nodes are ready, it will wait for up to 5 minutes for a node to become available.
func GetRandomNode(oc *exutil.CLI, role string) corev1.Node {
if node := getRandomNode(oc, role); isNodeReady(node) {
if node := getRandomNode(oc, role); IsNodeReady(node) {
return node
}

Expand All @@ -503,7 +503,7 @@ func GetRandomNode(oc *exutil.CLI, role string) corev1.Node {
framework.Logf("No ready nodes found with role '%s', waiting up to %s for a ready node to become available", role, waitPeriod)
var targetNode corev1.Node
o.Eventually(func() bool {
if node := getRandomNode(oc, role); isNodeReady(node) {
if node := getRandomNode(oc, role); IsNodeReady(node) {
targetNode = node
return true
}
Expand Down Expand Up @@ -549,7 +549,7 @@ func GetAllNodes(oc *exutil.CLI) ([]corev1.Node, error) {
}

// `isNodeReady` determines if a given node is ready
func isNodeReady(node corev1.Node) bool {
func IsNodeReady(node corev1.Node) bool {
// If the node is cordoned, it is not ready.
if node.Spec.Unschedulable {
return false
Expand Down Expand Up @@ -1038,7 +1038,7 @@ func GetNewReadyNodeInMachine(oc *exutil.CLI, machineName string) (corev1.Node,

// Check if node is in desiredStatus
framework.Logf("Checking if node '%v' is ready.", node.Name)
if isNodeReady(node) {
if IsNodeReady(node) {
framework.Logf("Node '%v' is ready.", node.Name)
desiredNode = node
err = nil
Expand Down