Alauda Container Platform GitOps Docs

Alauda Container Platform GitOps Sync and Health Status

Alauda Container Platform GitOps abstracts the state of Application resources by leveraging the status of underlying Kubernetes resources. The state of Application resources directly governs the state of associated ApplicationSet resources.

TOC

Sync Status ExplanationHealth Status ExplanationRecognition Rules

Sync Status Explanation

Both Kubernetes resources and applications have four sync states: Sync Failed, OutOfSync, Syncing, and Synced.

Sync StatusDescription
Sync FailedSynchronization failed due to network errors, configuration issues, or permissions problems. Check logs for root cause.
OutOfSyncCluster resource state diverges from Git-defined desired state. Manual/auto sync required.
SyncingActive reconciliation in progress between cluster state and Git-defined state.
SyncedCluster resource state matches Git-defined desired state.
INFO

Sync Status Display Priority: Priority order Sync Failed > OutOfSync > Syncing > Synced.

Examples:

  • If an Application has two resources with Syncing and Synced statuses, its overall status is Syncing.
  • If an ApplicationSet manages two Applications with Sync Failed and Synced statuses, its overall status is Sync Failed.

Health Status Explanation

Kubernetes resources and applications have six health states: Unknown, Missing, Degraded, Paused, Progressing, and Healthy.

Health StatusDescriptionReference Solution
UnknownUnable to determine health state, typically due to controller errors or missing status data.Inspect resource YAML's status.conditions for diagnostic details.
MissingResource not found in cluster.Initial creation: Wait for reconciliation
Accidental deletion: Trigger manual sync.
DegradedWorkload resources (e.g., Deployment) failed to achieve healthy state within timeout period (default: 10 mins).Investigate Pod failures (e.g., crashes, resource constraints).
PausedWorkload resources rollout intentionally paused (e.g., via kubectl rollout pause).Resume rollout if appropriate.
ProcessingResource created successfully but not fully ready (e.g., Pods initializing).Monitor until transition to Healthy/Degraded.
HealthyResource operating normally.-
INFO

Health Status Priority:Priority order Unknown > Missing > Degraded > Paused > Progressing > Healthy

Examples:

  • If an Application has resources with Healthy and Unknown statuses, its overall health is Unknown.
  • If an ApplicationSet manages Applications with Missing and Progressing statuses, its overall health is Missing.

Recognition Rules

Healthy status recognition rules for Kubernetes resources:

Resource TypeStatus
DeploymentRolling update completed with all replicas available.
StatefulSetUpdate completed with all pods ready.
ReplicaSetAll Pods healthy.
DaemonSetDesired number of Pods scheduled and healthy.
IngressLoadBalancer IP/hostname populated in status.
ServiceLoadBalancer IP/hostname populated (if applicable).
PVCStatus is Bound.
PodAll containers ready with no restarts exceeding thresholds.
JobJob completed successfully (.status.succeeded >= 1).
HPASuccessful scaling operation with current replicas matching desired count.