This message was deleted.
# harvester
a
This message was deleted.
a
Depending on how you are targeting your Harvester instance, maybe related to this: https://github.com/harvester/harvester/issues/2590
f
h
In my setup, I created a uplink bond between each Harvester node and external networks, as below
I also used iptables in the Harvester nodes. It took a very long time for the Harvester cluster to come up
and after that, there are a lot of unhealth events
Is my network setup adding a performance penalty on the cluster? As above screenshot, kube-api-server and calico-kube-controllers are not really ready.
f
You can try using a cni other than calico and not using the cloud provider and see if that works
441 Views