Description of problem: Desired state doesn't picks up the correct value and is getting NULL whic is invalid. {"level":"error","ts":1586351111.0703022,"logger":"enactmentconditions","msg":"Error notifying state NodeSelectorNotMatching","enactment":"host-172-16-0-22.nncp-default","error":"NodeNetworkConfigurationEnactment.nmstate.io \"host-172-16-0-22.nncp-default\" is invalid: status.desiredState: Invalid value: \"null\": status.desiredState in body must be of type object: $ oc get nncp NAME STATUS nncp-default-host-172-16-0-20 ConfigurationProgressing nncp-default-host-172-16-0-32 ConfigurationProgressing Version-Release number of selected component (if applicable): $ oc get csv -n openshift-cnv | awk ' { print $4 } ' | tail -n1 2.3.0 $ oc version Client Version: 4.4.0-0.nightly-2020-04-04-025830 Server Version: 4.4.0-rc.6 Kubernetes Version: v1.17.1 How reproducible: always Steps to Reproduce: 1. Setup a linux-bridge on worker nodes and make sure to set it up on primary NIC.(ens3) 2. Attached is the worker nodes file for bridge setup. 3. Make sure dhcp is set for only primary nic and is false for all secondary nic's attached. Actual results: Seeing issues during bridge setup and communication. Expected results: Bridge setup should work as expected on primary NIC and pods/vm's on worker1 nodes should be able to communicate with pods/vm's on worker2. Additional info:
Quique, would you please help me understand this issue? Is there a workaround? Should we expect it happening on production environment? What is causing it?
Have we test this with latest CNV version ? we added some fixes to kubernetes-nmstate before freezy last friday.
Knmstate is not rejecting the policy with a missing desiredState, looking at _types.go from policy it's not marked as optional so it's suppos to be requiered we have to investigate will be fixed at cnv 2.4 since this is not a blocker.
I have apply the policy at the bug at CNV 2.4 and we validating it as expected so we can close this bug. The NodeNetworkConfigurationPolicy "nncp-default-host" is invalid: spec.desiredState: Invalid value: "null": spec.desiredState in body must be of type object: "null"
Test Environment: Old versions: ============= $ oc get csv -n openshift-cnv | awk ' { print $4 } ' | tail -n1 2.3.0 1. Setup using config file mentioned in attachment. 2. Run oc apply. $ oc apply -f test.yaml nodenetworkconfigurationpolicy.nmstate.io/nncp-default-host created 3. Check status of nncp $ oc get nncp NAME STATUS nncp-default-host ConfigurationProgressing It is kind of hanged in this intermediate state. New versions: ============ $ oc get csv -n openshift-cnv | awk ' { print $4 } ' | tail -n1 2.4.0 $ oc apply -f test.yaml The NodeNetworkConfigurationPolicy "nncp-default-host" is invalid: spec.desiredState: Invalid value: "null": spec.desiredState in body must be of type object: "null" It is failing when we try to apply the policy so nncp is not getting created.
Please add 'fixed in version'
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/RHSA-2020:3194
The needinfo request[s] on this closed bug have been removed as they have been unresolved for 1000 days