The readiness probe for the DCAE Cloudify Manager is currently set to have a 1-second timeout. We have observed in at least one environment that the readiness-check.sh script can take over 2 seconds to execute. This causes the pod to marked as not ready.
Using kubectl edit to change the timeout (to 10 seconds) and the period (it was 10 seconds; changed to 30 seconds) allowed the readiness probe to complete successfully.
This change needs to be made permanent by adding the timeoutSeconds field to the deployment.yaml file for dcae-cloudify-manager and by updating the values.yaml file to include timeoutSeconds as a parameter and to change the existing value for periodSeconds.