DKP 2.6.1 Customer Incidents and Resolved Issues

The following resolved incidents are included in this release.

Prometheus has duplicate targets for the apiserver

D2IQ-98647

The default values in the kube-prometheus-stack chart included a duplicate scrape target for the Kubernetes API server. The duplicate scrape target is removed.

Running the DKP upgrade command twice causes a regression in the associated containerd secret

D2IQ-98947

In DKP 2.6, if a cluster was configured with a registry-mirror and credentials, running the update controlplane command twice corrupted the value of the containerd secret. This issue has been corrected.

Pre-provisioned Upgrade failed with Flatcar

An issue was resolved that caused Kubernetes upgrades to fail when using Flatcar OS with the pre-provisioned provider.

D2IQ-98979

Editing Credentials of a vSphere cluster in the DKP UI were not updated

An issue was resolved that could corrupt the configuration of vSphere clusters by editing their settings in the UI.

D2IQ-98762

Limits on Capi-controller needed raised

A performance problem was resolved after being observed to occur when creating large numbers of workload clusters.

D2IQ-99403

Insight alerts from an Essentials cluster converted into Managed land into the management cluster workspace

DKP resolved a problem where Insights were not displayed for Managed Clusters that had been expanded from Essential clusters.

D2IQ-98597

Grafana-Loki telemetry sending data back to Grafana

DKP disabled telemetry collection settings in the configuration of Grafana-Loki and Grafana.

D2IQ-99007

Increased Time to Attach Several Clusters in 2.6.0

In the 2.6.0 DKP release, when attaching clusters in a batch, there was an issue with increased time to attach several clusters. This issue has been resolved.

D2IQ-98664