Donato Bagarozza
Donato Bagarozza
@embik thanks for the information. We have a setup with 3 Master nodes and 6 worker nodes for 16 clusters. Master nodes has 16gb ram and 4 cores and worker...
as from time to time we also have dropouts in the metrics from user cluster:
we also have a another problem that prometheus alerts that apiserver will expire in some days. We opened an another issue: https://github.com/kubermatic/kubermatic/issues/9046 but kubeadm and the api prints the new...
metallb with nginx ingress controller, communication over nodeport service. We use as provider vsphere with ubuntu template 18.04.6
@embik we have tried some things..direct kubectl forward the apiserver from kkp user cluster, the request works. Forward via the nodeport-service only partially, when we reboot the nodeport-proxy-envoy pods it...
we updated kubermatic to 2.17.5 but the problem still exist. We want to allocate more ram at the nodeport proxy envoy, but this does not work. We updated our seedconfig,...