The issue blocks all sts related cluster testing, add TestBlocker keywords

Does this only apply to AWS? If so, this will be fixed by PR: https://github.com/openshift/cloud-network-config-controller/pull/13

Now we support token credentials only on aws and gcp, only haves issue for aws platform, will verify it with the fix PR

Verified with cluster-bot image with the fix PR merged, the crash issue has fixed, but the operator hits another issue

####
caused by: WebIdentityErr: unable to read file at /var/run/secrets/openshift/serviceaccount/token
caused by: open /var/run/secrets/openshift/serviceaccount/token: no such file or directory, requeuing in node workqueue
E0104 02:53:03.460524       1 controller.go:165] error syncing 'ip-10-0-169-247.us-east-2.compute.internal': error retrieving the private IP configuration for node: ip-10-0-169-247.us-east-2.compute.internal, err: error: cannot list ec2 instance for node: ip-10-0-169-247.us-east-2.compute.internal, err: WebIdentityErr: failed fetching WebIdentity token: 
caused by: WebIdentityErr: unable to read file at /var/run/secrets/openshift/serviceaccount/token
caused by: open /var/run/secrets/openshift/serviceaccount/token: no such file or directory, requeuing in node workqueue

@jdiaz is the developer of Cloud Credential Operator , I think he can give some information if you need help.

Based on the error message, it looks like the projected ServiceAccount token has not been mounted into /var/run/secrets/openshift/serviceaccount/token.

Looking at the github repo, I didn't see where the Deployment manifest is defined, but I would expect the Deployment that is defined for running this 'cloud-network-config-controller' software to have a volume mount that looks like https://github.com/openshift/cluster-image-registry-operator/blob/master/manifests/07-operator.yaml#L75-L77

The Hypershift fix also includes PR: https://github.com/openshift/cluster-network-operator/pull/1268/files which loads the project service account token. I suspect that should fix the problem completely and maybe we should retest once that PR merges.

I saw another error logs, please check
###
E0113 01:50:55.536951       1 controller.go:165] error syncing 'ip-10-0-184-222.us-east-2.compute.internal': error retrieving the private IP configuration for node: ip-10-0-184-222.us-east-2.compute.internal, err: error: cannot list ec2 instance for node: ip-10-0-184-222.us-east-2.compute.internal, err: WebIdentityErr: failed to retrieve credentials
caused by: AccessDenied: Not authorized to perform sts:AssumeRoleWithWebIdentity
	status code: 403, request id: 3572c767-04c8-42c0-86c1-ba0f2c5d5091, requeuing in node workqueue
E0113 01:50:55.554710       1 controller.go:165] error syncing 'ip-10-0-202-212.us-east-2.compute.internal': error retrieving the private IP configuration for node: ip-10-0-202-212.us-east-2.compute.internal, err: error: cannot list ec2 instance for node: ip-10-0-202-212.us-east-2.compute.internal, err: WebIdentityErr: failed to retrieve credentials
caused by: AccessDenied: Not authorized to perform sts:AssumeRoleWithWebIdentity
	status code: 403, request id: 896bdc0d-23dd-44fa-be42-52e5a4bf7893, requeuing in node workqueue
I0113 01:53:39.559676       1 controller.go:160] Dropping key 'ip-10-0-143-215.us-east-2.compute.internal' from the node workqueue
I0113 01:53:39.578314       1 controller.go:160] Dropping key 'ip-10-0-202-212.us-east-2.compute.internal' from the node workqueue
I0113 01:53:39.578350       1 controller.go:160] Dropping key 'ip-10-0-191-236.us-east-2.compute.internal' from the node workqueue
I0113 01:53:39.578356       1 controller.go:160] Dropping key 'ip-10-0-198-79.us-east-2.compute.internal' from the node workqueue
I0113 01:53:39.578363       1 controller.go:160] Dropping key 'ip-10-0-184-222.us-east-2.compute.internal' from the node workqueue
I0113 01:53:39.578365       1 controller.go:160] Dropping key 'ip-10-0-145-35.us-east-2.compute.internal' from the node workqueue

The CredentialsRequests I used for the cloud-network-config-controller is
cat > "${creds_dir}/0000_50_cloud-network_00_credentials-request.yaml" <<EOF
---
apiVersion: cloudcredential.openshift.io/v1
kind: CredentialsRequest
metadata:
  name: openshift-cloud-network-config-controller
  namespace: openshift-cloud-credential-operator
spec:
  providerSpec:
    apiVersion: cloudcredential.openshift.io/v1
    kind: AWSProviderSpec
    statementEntries:
    - action:
      - ec2:DescribeInstances
      - ec2:DescribeInstanceStatus
      - ec2:DescribeInstanceTypes
      - ec2:UnassignPrivateIpAddresses
      - ec2:AssignPrivateIpAddresses
      - ec2:UnassignIpv6Addresses
      - ec2:AssignIpv6Addresses
      - ec2:DescribeSubnets
      - ec2:DescribeNetworkInterfaces
      effect: Allow
      resource: '*'
  secretRef:
    name: cloud-credentials
    namespace: openshift-cloud-network-config-controller
  serviceAccountNames:
  - cloud-network-config-controllers
EOF

attanched must-gather logs

@Casey

Could you please have a look at this bug? Not sure how STS credentials should work and what might be missing. PR: https://github.com/openshift/cluster-network-operator/pull/1268/files mentioned that AssumeRoleWithWebIdentity should work on AWS, but it seems it doesn't

Tested using payload 4.10.0-0.nightly-2022-01-17-182202 with PR 1275 merged, the same issuse

@jdiaz Could you help take a look? I checked s3/key.json is:
###
{
    "keys": [
        {
            "use": "sig",
            "kty": "RSA",
            "kid": "0zCe_XCNey2bBDh9fYuVVCitvEYP7bbKgFb_jOpuTEU",
            "alg": "RS256",
            "n": "uN0XAwS4shiBQinahpsbrHOCgAT4jssPWjEDs2jerBOGPETySxDlc295HkuZCb8tVLZdu3PKQlxXDEIzBuTO1DW2yWTHONVu7tnEywtuOJVgnNTu-95LU1oIFLZMnxjVmuaZSygTgDQI_p_K9wmiDmQvGudJUHVokd0N0rM63Qf1JE1UmDOgdS6jfHbDKgBLI3lZfMi7Xfdb9YMQ8JG8fIh86yB6iPWPyasNqwSAUiIoTPdnjMm6s2mFM-AVMof3z7Fs4MThFRpyNkQUKfxhBOZRsLwgoK1H0Z1N9F5VAcRTC6tIkBFYY_ty74KQtzTy1neZNam8mRVPavjGdnUbse9J3Fn7kS5d9iylGg-2WDFNnrjwUTU3Cqh8BGkFV5UubAWLjald-0-yJUEC4EuBGH0sJkYnVHTONUNFmyO4F1nbhwISZQNtjayROKDRRdSYuEkC1KKWYZiNbmn8bUCjDqD3f5Sy_D0DQEqdLbv9t5Bry0stugwwjmdS_NamMSVpZtCfVcP0B0seqMU8kIhDEE_sYQpG72NgLLgg1JZEMV-8FafoL2z3_uJ3ZEJc7ECc-nM_TEDfEPObEO6orZhK87tXfPr4bomAMZGchSEE3lZhaj5y61iRYtUPZ6aYR3n5TZKY9PRzX6Z0-oABt_bYCJsJ80bw5pkeP8Q5XOoVWF8",
            "e": "AQAB"
        }
    ]
}

Token for CCNC is:
###
$ oc rsh cloud-network-config-controller-7dbdf9bb85-4xct7
sh-4.4$ cat /var/run/secrets/openshift/serviceaccount/token | awk -F. '{ print $1 }' | base64 -d
{"alg":"RS256","kid":"0zCe_XCNey2bBDh9fYuVVCitvEYP7bbKgFb_jOpuTEU"}base64: invalid input
sh-4.4$ cat /var/run/secrets/openshift/serviceaccount/token | awk -F. '{ print $2 }' | base64 -d
{"aud":["openshift"],"exp":1642478458,"iat":1642474858,"iss":"https://lwansts0118-14958-oidc.s3.us-east-2.amazonaws.com","kubernetes.io":{"namespace":"openshift-cloud-network-config-controller","pod":{"name":"cloud-network-config-controller-7dbdf9bb85-4xct7","uid":"6f9dfb1b-f9b5-480b-ae80-70cca4b87ecc"},"serviceaccount":{"name":"cloud-network-config-controller","uid":"4b630e33-3e18-45ed-bd77-5e07f94573c5"}},"nbf":1642474858,"sub":"system:serviceaccount:openshift-cloud-network-config-controller:cloud-network-config-controller"}


I can't see what issue's on it, but the pod show this error:
###
E0118 03:08:20.806257       1 controller.go:165] error syncing 'ip-10-0-213-232.us-east-2.compute.internal': error retrieving the private IP configuration for node: ip-10-0-213-232.us-east-2.compute.internal, err: error: cannot list ec2 instance for node: ip-10-0-213-232.us-east-2.compute.internal, err: WebIdentityErr: failed to retrieve credentials
caused by: AccessDenied: Not authorized to perform sts:AssumeRoleWithWebIdentity
	status code: 403, request id: a7f6aa50-fb30-40cc-99b9-ed92c6497aee, requeuing in node workqueue

also Trust Relationship:

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Principal": {
        "Federated": "arn:aws:iam::301721915996:oidc-provider/lwansts0118-14958-oidc.s3.us-east-2.amazonaws.com"
      },
      "Action": "sts:AssumeRoleWithWebIdentity",
      "Condition": {
        "StringEquals": {
          "lwansts0118-14958-oidc.s3.us-east-2.amazonaws.com:sub": "system:serviceaccount:openshift-cloud-network-config-controller:cloud-network-config-controllers"
        }
      }
    }
  ]
}

@cdc GCP workload identity cluster has the similar issue, the installation can successful, so I didn't notice the issue before. Do you want to use one bug to track or need me to create another bug for GCP platform?

###
E0118 05:22:55.659257       1 controller.go:165] error syncing 'lwanstsg0118-5lwqc-worker-c-b2g8k.c.openshift-qe.internal': error retrieving the private IP configuration for node: lwanstsg0118-5lwqc-worker-c-b2g8k.c.openshift-qe.internal, err: error retrieving instance associated with node, err: Get "https://compute.googleapis.com/compute/v1/projects/openshift-qe/zones/us-central1-c/instances/lwanstsg0118-5lwqc-worker-c-b2g8k?alt=json&prettyPrint=false": oauth2/google: status code 403: {
  "error": {
    "code": 403,
    "message": "The caller does not have permission",
    "status": "PERMISSION_DENIED"
  }
}
, requeuing in node workqueue

No, PR 1283 can't fix the issue, when I create sts cluster,I already manually add serviceAccountNames field to CredentialsRequest, I use ccoctl tool to create sts related manifests follow this offical doc https://docs.openshift.com/container-platform/4.9/authentication/managing_cloud_provider_credentials/cco-mode-sts.html#sts-mode-installing ,

if there is no serviceAccountNames filed, ccoctl tool will hit an issue like :
#####
2021/12/07 11:02:07 Failed to process IAM Roles: Failed while processing each CredentialsRequest: error while creating Role policy document for openshift-cluster-api-aws: CredentialsRequest must provide ServieAccounts to bind the Role policy to
#####

so I have manually added serviceAccountNames before I launch cluster

@jdiaz , sorry for the trouble above, I checked with Casey, I gave serviceAccountNames "cloud-network-config-controllers" value, but it should be cloud-network-config-controller without 's', please ignore above comments.

The permission issue has fixed, will move the bug to verified

####pod logs####
I0118 11:34:35.683463       1 leaderelection.go:258] successfully acquired lease openshift-cloud-network-config-controller/cloud-network-config-controller-lock
I0118 11:34:35.684267       1 controller.go:88] Starting node controller
I0118 11:34:35.684278       1 controller.go:91] Waiting for informer caches to sync for node workqueue
I0118 11:34:35.684761       1 controller.go:88] Starting secret controller
I0118 11:34:35.684771       1 controller.go:91] Waiting for informer caches to sync for secret workqueue
I0118 11:34:35.684780       1 controller.go:88] Starting cloud-private-ip-config controller
I0118 11:34:35.684786       1 controller.go:91] Waiting for informer caches to sync for cloud-private-ip-config workqueue
I0118 11:34:35.687749       1 controller.go:182] Assigning key: ip-10-0-194-96.us-east-2.compute.internal to node workqueue
I0118 11:34:35.688394       1 controller.go:182] Assigning key: ip-10-0-141-229.us-east-2.compute.internal to node workqueue
I0118 11:34:35.688412       1 controller.go:182] Assigning key: ip-10-0-186-235.us-east-2.compute.internal to node workqueue
I0118 11:34:35.784845       1 controller.go:96] Starting cloud-private-ip-config workers
I0118 11:34:35.784862       1 controller.go:96] Starting node workers
I0118 11:34:35.784872       1 controller.go:102] Started cloud-private-ip-config workers
I0118 11:34:35.784880       1 controller.go:102] Started node workers
I0118 11:34:35.784896       1 controller.go:96] Starting secret workers
I0118 11:34:35.784922       1 controller.go:102] Started secret workers

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 (Moderate: OpenShift Container Platform 4.10.3 security update), 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-2022:0056