Description of problem: When trying to create a cluster with, - machineNetwork: 10.0.0.0/16 - only one AZ ca-central-1b the installer creates subnets with 10.0.0.0/20 and 10.0.128.0/20. With this setup only 4096*2 machines can be created. and the rest of the VPC range is left completely used. This is esp problematic for smaller machineNetworks. The installer should have created the subnets with 10.0.0.0/17 and 10.0.128.0/17 ranges. example of the install-config.yaml used. ``` apiVersion: v1 baseDomain: XXXXXX controlPlane: name: master replicas: 3 platform: aws: zones: - ca-central-1b compute: - name: worker replicas: 3 platform: aws: zones: - ca-central-1b metadata: name: xxxxx networking: machineNetwork: - cidr: 10.0.0.0/16 platform: aws: region: ca-central-1 ```
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2020:2409
Just capturing results from my test run just now with 4.7 fc4: Started with compute: - architecture: amd64 hyperthreading: Enabled name: worker platform: aws: zones: - us-east-2b replicas: 3 controlPlane: architecture: amd64 hyperthreading: Enabled name: master platform: aws: zones: - us-east-2b replicas: 3 ... networking: clusterNetwork: - cidr: 10.128.0.0/14 hostPrefix: 23 machineNetwork: - cidr: 10.0.0.0/16 networkType: OpenShiftSDN serviceNetwork: - 172.30.0.0/16 platform: aws: region: us-east-2 Ends with two subnets: tszeaws012921b-674lr-public-us-east-2b 10.0.0.0/17 tszeaws012921b-674lr-private-us-east-2b 10.0.128.0/17 It also doesn't apply to GCP as GCP uses one subnet across all zones (AWS needs different subnets for each zone).