upgrade from 1.6 to 1.7: longhorn-ready unit times out because the new instance manager pod cannot be scheduled due to insufficient cpu
Job #15797393125 failed for 3069c3c6:
in the 1.6 to 1.7 upgrade pipelines, update-workload-cluster fails:
Unit timeout exceeded: unit Kustomization/longhorn-ready did not become ready after 5m0s
└─ timeout waiting for: [Job/rke2-capm3-virt/longhorn-ready status: 'InProgress']the new instance manager pod stays in error:
OutOfcpu: Node didn't have enough resource: cpu, requested: 480, used: 3621, capacity: 4000
by default Instance Manager pods reserve 12% of the total CPU capacity on a node, we might need to reconsider decreasing this reservation