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
Problem
When using AWS EKS as a Kubernetes platform and with Calico as the CNI, we are required to run certain pods on the host network. There is currently no way to override the default webhook port values for each Deployment when using the Knative operator to create the KnativeServing and KnativeEventing resources. As a result, we must configure taints and tolerations for our pods and nodes so that the host network pods are scheduled on their own individual nodes, which removes the conflict of needing to run on a specific port, but introduces additional cost and management for the dedicated EC2 instances that are running the pods. A specific example of this when using EKS + Calico is the KnativeEventing eventing-webhook and imc-controller pods needing to be run with hostNetwork: true and only one Node available.
Persona:
Which persona is this feature for?
System Operator
Exit Criteria
A measurable (binary) test that would indicate that the problem has been resolved.
Override the default webhook port for KnativeEventing deployments and testing to see if the eventing-webhook and imc-controller pods can be run on the same Node when hostNetwork: true
Time Estimate (optional):
How many developer-days do you think this may take to resolve?
Additional context (optional)
Add any other context about the feature request here.
The text was updated successfully, but these errors were encountered:
Problem
When using AWS EKS as a Kubernetes platform and with Calico as the CNI, we are required to run certain pods on the host network. There is currently no way to override the default webhook port values for each Deployment when using the Knative operator to create the KnativeServing and KnativeEventing resources. As a result, we must configure taints and tolerations for our pods and nodes so that the host network pods are scheduled on their own individual nodes, which removes the conflict of needing to run on a specific port, but introduces additional cost and management for the dedicated EC2 instances that are running the pods. A specific example of this when using EKS + Calico is the KnativeEventing
eventing-webhook
andimc-controller
pods needing to be run withhostNetwork: true
and only one Node available.Persona:
Which persona is this feature for?
System Operator
Exit Criteria
A measurable (binary) test that would indicate that the problem has been resolved.
Override the default webhook port for KnativeEventing deployments and testing to see if the
eventing-webhook
andimc-controller
pods can be run on the same Node whenhostNetwork: true
Time Estimate (optional):
How many developer-days do you think this may take to resolve?
Additional context (optional)
Add any other context about the feature request here.
The text was updated successfully, but these errors were encountered: