You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
You can fix this by marking your infra with the label 'node-role.kubernetes.io/compute=true', as well. This is because we need to deploy our Nuage CNI, VRS and Infra pod (which are deployed as part of a daemonset) on the infra nodes as well. We've noticed there are improvements possible that might not require that label in the future, but for now, this is required.
Very valid remark on documentation. I'll be adding this as an extra note.
@pdellaert I absolutely tried that. The nodes did get labelled properly, however, the Nuage Infra PODS, OCP router and registry would not begin running. I'd like to reproduce this and we can have a look. Definitely something to do with the labelling/node group structure.
Tried many deployments with the following configuration and labels for nodes:
3 x Masters
2 x Infra
3 x compute
Installation would fail as registry and router pods would not not create when assigned and scheduled to infra roles, need to be assigned to master.
Here are the node group labels used:
Can documentation/option be updated to support this typical OpenShift deployment architecture?
The text was updated successfully, but these errors were encountered: