Skip to content

core: fix ceph health check up status check #16595

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
travisn merged 2 commits into from
Oct 14, 2025
Merged

core: fix ceph health check up status check #16595

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
454c221
core: fix cep health check up status check
parth-gr Oct 10, 2025
3f5ebe2
external: add cluster connected condition in case of external mode
parth-gr Oct 14, 2025
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
29 changes: 7 additions & 22 deletions pkg/operator/ceph/cluster/cephstatus.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -111,14 +111,6 @@ func (c *cephStatusChecker) checkStatus(ctx context.Context) {
var err error

logger.Debugf("checking health of cluster")

condition := cephv1.ConditionReady
reason := cephv1.ClusterCreatedReason
if c.isExternal {
condition = cephv1.ConditionConnected
reason = cephv1.ClusterConnectedReason
}

// Check ceph's status
status, err = cephclient.StatusWithUser(c.context, c.clusterInfo)
if err != nil {
Expand All @@ -127,22 +119,13 @@ func (c *cephStatusChecker) checkStatus(ctx context.Context) {
return
}
logger.Errorf("failed to get ceph status. %v", err)

message := "Failed to configure ceph cluster"
if c.isExternal {
message = "Failed to configure external ceph cluster"
}
status := cephStatusOnError(err.Error())
c.updateCephStatus(status, condition, reason, message, v1.ConditionFalse)
c.updateCephStatus(status)
return
}

logger.Debugf("cluster status: %+v", status)
message := "Cluster created successfully"
if c.isExternal {
message = "Cluster connected successfully"
}
c.updateCephStatus(&status, condition, reason, message, v1.ConditionTrue)
c.updateCephStatus(&status)

if status.Health.Status != "HEALTH_OK" {
logger.Debug("checking for stuck pods on not ready nodes")
Expand Down Expand Up @@ -172,7 +155,7 @@ func (c *cephStatusChecker) configureHealthSettings(status cephclient.CephStatus
}

// updateCephStatus updates an object with a given status
func (c *cephStatusChecker) updateCephStatus(status *cephclient.CephStatus, condition cephv1.ConditionType, reason cephv1.ConditionReason, message string, conditionStatus v1.ConditionStatus) {
func (c *cephStatusChecker) updateCephStatus(status *cephclient.CephStatus) {
clusterName := c.clusterInfo.NamespacedName()
cephCluster, err := c.context.RookClientset.CephV1().CephClusters(clusterName.Namespace).Get(c.clusterInfo.Context, clusterName.Name, metav1.GetOptions{})
if err != nil {
Expand All @@ -197,8 +180,10 @@ func (c *cephStatusChecker) updateCephStatus(status *cephclient.CephStatus, cond
}

// Update condition
logger.Debugf("updating ceph cluster %q status and condition to %+v, %v, %s, %s", clusterName.Namespace, status, conditionStatus, reason, message)
opcontroller.UpdateClusterCondition(c.context, cephCluster, c.clusterInfo.NamespacedName(), k8sutil.ObservedGenerationNotAvailable, condition, conditionStatus, reason, message, true)
logger.Debugf("updating ceph cluster %q status to %+v", clusterName.Namespace, status)
if err := reporting.UpdateStatus(c.context.Client, cephCluster); err != nil {
logger.Errorf("failed to update cluster status to %+v. %v", status, err)
}
}

// toCustomResourceStatus converts the ceph status to the struct expected for the CephCluster CR status
Expand Down
2 changes: 1 addition & 1 deletion pkg/operator/ceph/cluster/cluster_external.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -188,7 +188,7 @@ func (c *ClusterController) configureExternalCephCluster(cluster *cluster) error
}
}

// We don't update the connection status since it is done by the health go routine
opcontroller.UpdateCondition(c.OpManagerCtx, c.context, c.namespacedName, k8sutil.ObservedGenerationNotAvailable, cephv1.ConditionConnected, v1.ConditionTrue, cephv1.ClusterConnectedReason, "Cluster connected successfully")
return nil
}

Expand Down
Loading