Description of problem: We should have the ability to read port for reaching vCenter from provider config. https://vmware.github.io/vsphere-storage-for-kubernetes/documentation/k8s-secret.html#create-a-k8s-secret https://vmware.github.io/vsphere-storage-for-kubernetes/documentation/existing.html Steps for QE: 1. oc edit cm cloud-provider-config -n openshift-config 2. Add port under Global section [Global] port = "443"
Cluster version is 4.5.0-0.nightly-2020-06-01-043833 Steps : 1.Edited the cm successfully ( oc edit cm cloud-provider-config -n openshift-config) . . . apiVersion: v1 data: config: | [Global] secret-name = "vsphere-creds" secret-namespace = "kube-system" insecure-flag = "1" port = "443" . . Step 2. Created a new machineset and scaled [miyadav@miyadav ManualRun]$ oc get machineset NAME DESIRED CURRENT READY AVAILABLE AGE miyadav-0601-jgfsb-worker 2 2 2 2 54m miyadav-0601-jgfsb-worker-new 1 1 1 1 7m30s [miyadav@miyadav ManualRun]$ oc get machines -o wide NAME PHASE TYPE REGION ZONE AGE NODE PROVIDERID STATE miyadav-0601-jgfsb-master-0 Running 54m miyadav-0601-jgfsb-master-0 vsphere://420b9520-1c09-fbbc-bfde-215172b6b6b3 poweredOn miyadav-0601-jgfsb-master-1 Running 54m miyadav-0601-jgfsb-master-1 vsphere://420b18d9-23e8-5712-00de-53564336f755 poweredOn miyadav-0601-jgfsb-master-2 Running 54m miyadav-0601-jgfsb-master-2 vsphere://420bddcd-7a29-0d86-4ebc-432f02346a14 poweredOn miyadav-0601-jgfsb-worker-9cw9m Running 46m miyadav-0601-jgfsb-worker-9cw9m vsphere://420be8f0-ef7a-6078-90f1-7d0021f96202 poweredOn miyadav-0601-jgfsb-worker-new-2wdpn Running 7m37s miyadav-0601-jgfsb-worker-new-2wdpn vsphere://420be039-0fd8-8d48-ebae-631a486fdba4 poweredOn miyadav-0601-jgfsb-worker-tjqct Running 46m miyadav-0601-jgfsb-worker-tjqct vsphere://420b48ce-178c-eef0-3ac2-be325f7d3a69 poweredOn Actual & Expected : Machinescaling works fine , with or without port config ( creating deleting machines , machineset looks good) Additional Info : I observed below please have a look - [miyadav@miyadav ManualRun]$ oc get pods NAME READY STATUS RESTARTS AGE cluster-autoscaler-operator-6cddc8fc88-n7xv5 2/2 Running 0 2m45s machine-api-controllers-845c987f7b-lm2dc 4/4 Running 0 2m45s machine-api-operator-6d66bcbffc-9t44h 2/2 Running 0 2m45s scaled down machineset [miyadav@miyadav ManualRun]$ oc scale machineset miyadav-0601-jgfsb-worker-new --replicas=1 machineset.machine.openshift.io/miyadav-0601-jgfsb-worker-new scaled All pods of the openshift-machine-api namespace gets recreated [miyadav@miyadav ManualRun]$ oc get pods NAME READY STATUS RESTARTS AGE cluster-autoscaler-operator-6cddc8fc88-hpcjz 0/2 ContainerCreating 0 8s cluster-autoscaler-operator-6cddc8fc88-n7xv5 0/2 Terminating 0 4m23s machine-api-controllers-845c987f7b-kzst7 0/4 ContainerCreating 0 7s machine-api-controllers-845c987f7b-lm2dc 0/4 Terminating 0 4m23s machine-api-operator-6d66bcbffc-9t44h 0/2 Terminating 0 4m23s machine-api-operator-6d66bcbffc-rlgf9 0/2 ContainerCreating 0 8s This doesnt happen when we do not have port 443 in provider config(scale up and down happens and same pod is present ) , it happens only when we scale down for provider config having 443 port When ever this happens we get below logs : . . . I0601 09:12:01.846947 1 streamwatcher.go:114] Unexpected EOF during watch stream event decoding: unexpected EOF I0601 09:12:01.847502 1 streamwatcher.go:114] Unexpected EOF during watch stream event decoding: unexpected EOF error: unexpected EOF . .
After checking with Alex , they are internal k8s errors and not related to machine-api , moving this PR as VERIFIED.
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