All workloads on the faulty node have now started up on a new node.
Posted Feb 13, 2026 - 12:46 UTC
Identified
The node is cordoned off and now marked as unhealthy. Once a new node has been started up, we will start draining all the workloads on this node. The root cause is the kubelet having problems with managing Pod lifecycle, in particular termination, so if draining adds to the problem, all workloads on that node will be forcefully terminated. Downtime can be expected.
Posted Feb 13, 2026 - 12:37 UTC
Investigating
We are currently investigating this issue.
Posted Feb 13, 2026 - 12:35 UTC
This incident affected: Kubernetes critical components (Kubernetes [Hellman] - Capacity/Scheduling).