-
Feature Request
-
Resolution: Unresolved
-
Major
-
None
-
None
-
False
-
None
-
False
-
Not Selected
-
-
-
1. Proposed title of this feature request
ScaledObject status should provide more details to improve troubleshooting experience
2. What is the nature and description of the request?
Custom Metrics Autoscaler is a great tool, showing good adoption in the field. It though is missing out detailed information in the ScaledObject resource .status when something is not working as expected. For troubleshooting purpose, it's often required to access logs from the components in openshift-keda which though is restricted in enterprise environment.
To improve overall developer and user experience, it's requested to improve the .status messaging in ScaledObject to better highlight why the scaler may not work as expected.
status:
conditions:
- message: failed to ensure HPA is correctly created for ScaledObject
reason: ScaledObjectCheckFailed
status: "False"
type: Ready
- message: ScaledObject check failed
reason: UnknownState
status: Unknown
type: Active
- status: Unknown
type: Fallback
- status: Unknown
type: Paused
originalReplicaCount: 1
scaleTargetGVKR:
group: apps
kind: Deployment
resource: deployments
version: v1
scaleTargetKind: apps/v1.Deployment
The above shows an example where the secret in TriggerAuthentication is invalid. But based on the message this is not clear and therefore difficult to troubleshoot for a regular user.
2024-06-03T10:42:15Z ERROR Failed to create new HPA resource {"controller": "scaledobject", "controllerGroup": "keda.sh", "controllerKind": "ScaledObject", "ScaledObject": {"name":"quarkus-app-scaledobject","namespace":"project-100"}, "namespace": "project-100", "name": "quarkus-app-scaledobject", "reconcileID": "16572472-af58-46fc-a825-07b3cbf33168", "HPA.Namespace": "project-100", "HPA.Name": "keda-hpa-quarkus-app-scaledobject", "error": "error parsing prometheus metadata: no bearer token provided"} github.com/kedacore/keda/v2/controllers/keda.(*ScaledObjectReconciler).createAndDeployNewHPA /remote-source/keda/app/controllers/keda/hpa.go:52 github.com/kedacore/keda/v2/controllers/keda.(*ScaledObjectReconciler).ensureHPAForScaledObjectExists /remote-source/keda/app/controllers/keda/scaledobject_controller.go:438 github.com/kedacore/keda/v2/controllers/keda.(*ScaledObjectReconciler).reconcileScaledObject /remote-source/keda/app/controllers/keda/scaledobject_controller.go:267 github.com/kedacore/keda/v2/controllers/keda.(*ScaledObjectReconciler).Reconcile /remote-source/keda/app/controllers/keda/scaledobject_controller.go:180 sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Reconcile /remote-source/keda/app/vendor/sigs.k8s.io/controller-runtime/pkg/internal/controller/controller.go:119 sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).reconcileHandler /remote-source/keda/app/vendor/sigs.k8s.io/controller-runtime/pkg/internal/controller/controller.go:316 sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).processNextWorkItem /remote-source/keda/app/vendor/sigs.k8s.io/controller-runtime/pkg/internal/controller/controller.go:266 sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Start.func2.2 /remote-source/keda/app/vendor/sigs.k8s.io/controller-runtime/pkg/internal/controller/controller.go:227
The above shows the logs from the keda Operator providing definitely more details where to look and what potentially could be wrong. Having such kind of information in the .status of the ScaledObject or emitted as Event would greatly improve the experience for developers using the Custom Metrics Autoscaler.
3. Why does the customer need this? (List the business requirements here)
In enterprise OpenShift Container Platform 4 environment, consumers are usually restricted to their own namespace and don't have access to openshift-keda where the respective controller and hence data are available, showing my certain ScaledObject may not work as expected.
It's therefore required to provide the required information about the state of the ScaledObject in the respective ScaledObject resource or via event in the namespace the ScaledObject is located so that consumers can troubleshoot and fix issues on their own, rather than having to work with the cluster-admin group.
4. List any affected packages or components.
Custom Metrics Autoscaler