Investigate Missing Kubernetes Warning Events in OneUptime #5179

Describe the bug
Kubernetes warning events are visible in the Kubernetes environment, but we need to investigate why these events are not visible or surfaced as expected in OneUptime.

The attached screenshot shows the following warning events:

OOMKilled: Memory cgroup out-of-memory events indicating that processes were killed.

FailedGetResourceMetric: The Horizontal Pod Autoscaler (HPA) cannot retrieve CPU utilization metrics because the resource metrics API returns no metrics.

FailedComputeMetricsReplicas: HPA cannot calculate the desired replica count because the CPU resource metric is unavailable or invalid.

The affected objects include AKS nodes and HPAs in the aks-istio-system and app-routing-system namespaces.

Expected Behavior

OneUptime should make relevant Kubernetes warning events visible and accessible for investigation when the corresponding integration supports collecting them.

Users should be able to identify event details, including the message, severity, reason, source, affected object, namespace, and occurrence count.

Actual Behavior

The events are present in the Kubernetes event view shown in the attached screenshot, but we need OneUptime to investigate why they were not visible or surfaced as expected in its interface.

The point of this issue is to determine where the events are being lost or excluded, rather than assume that they were never collected.

Screenshots

Desktop (please complete the following information):

• OS: [Windows]
• Browser [Chrome]
• Version [155.0.8059.85 ]

Deployment Type
Self Hosted 14.0.80