Description of problem: cluster is inaccessible with kas containers showing many "tls: bad certificate" and "x509: certificate signed by unknown authority" logs Version-Release number of selected component (if applicable): 4.5.0-0.nightly-2020-08-07-024812 rhcos_ami: "rhcos-45.82.202008010929-0" How reproducible: always Steps to Reproduce: 1. Install upi vsphere (6.7) env with above version Actual results: 1. Installation failed. The installation logs ended with: ... The connection to the server api.jima45upitest01.qe.devcluster.openshift.com:6443 was refused - did you specify the right host or port? Expect 3 ready master nodes, but found 0, waiting... Try 1, before approve csr, run 'oc get node --no-headers -l 'node-role.kubernetes.io/master'; oc get csr' The connection to the server api.jima45upitest01.qe.devcluster.openshift.com:6443 was refused - did you specify the right host or port? The connection to the server api.jima45upitest01.qe.devcluster.openshift.com:6443 was refused - did you specify the right host or port? Trying to approve CSR requests for nodes...... The connection to the server api.jima45upitest01.qe.devcluster.openshift.com:6443 was refused - did you specify the right host or port? ... Expected results: 1. Should succeed. Additional info: ssh to master, check kas containers: [root@control-plane-2 xxia-debug]# crictl ps -a | grep kube-apiserver f22de68ab6af5 ccd7b0e07b19bd94230eb49b8da93821461cc3aecc1e38afe00136535adf6de2 4 hours ago Running kube-apiserver-insecure-readyz 0 5baaa11702884 73cc851b3ee8e ccd7b0e07b19bd94230eb49b8da93821461cc3aecc1e38afe00136535adf6de2 4 hours ago Running kube-apiserver-cert-regeneration-controller 0 5baaa11702884 857a95b044166 ccd7b0e07b19bd94230eb49b8da93821461cc3aecc1e38afe00136535adf6de2 4 hours ago Running kube-apiserver-cert-syncer 0 5baaa11702884 81b4b51d7d341 7a8f32d752782629e9d97f69a619b1570377011a0fb544a5e7fceced81edf281 4 hours ago Running kube-apiserver Check kube-apiserver logs (see attachments): crictl logs 81b4b51d7d341 &> kas.log crictl logs 857a95b044166 &> kube-apiserver-cert-syncer.log kas.log shows many "I0807 07:18:08.429523 1 log.go:172] http: TLS handshake error from [::1]:56428: remote error: tls: bad certificate" and 'v1.xxxxxx.openshift.io" failed with: failed to retrieve openAPI spec, http error: ResponseCode: 503, Body: service unavailable' kube-apiserver-cert-syncer.log shows many 'Failed to list *v1.xxxx: Get https://localhost:6443/api/v1/namespaces/openshift-kube-apiserver/xxxx?limit=500&resourceVersion=0: x509: certificate signed by unknown authority'
This issue is happened when using new rhcos template(rhcos-45.82.202008010929-0) to boot up vms It works well with previous rhcos tmplate(rhcos-45.82.202007141718-0).
Issue is identified due to Bug 1870038, after applying workaround in Bug 1870038, installation is successful, this issue is not produced any more. So close this bug.
*** This bug has been marked as a duplicate of bug 1870038 ***