tiraboschi
commented
2 weeks ago @kgfathur is this still happening once you create the CR for the HCO operator?
Closed kgfathur closed 1 week ago
tiraboschi
commented
2 weeks ago @kgfathur is this still happening once you create the CR for the HCO operator?
kgfathur
commented
2 weeks ago @kgfathur is this still happening once you create the CR for the HCO operator?
it seems that I missed some steps. I don't have it on my cluster:
$ kubectl get hco -A
No resources foundDo you mean the CR for the HyperConverged, like on this example hco.cr.yaml
Currently, before creating the CR for the HCO, the ssp-operator deployment not failing to CrashLoopBackOff.
$ kubectl get pod -n operators
NAME READY STATUS RESTARTS AGE
cdi-operator-7d9ff47dcc-t9ms6 1/1 Running 0 3h3m
cluster-network-addons-operator-5b54878c48-mr89l 2/2 Running 0 3h1m
hco-operator-7fd5bd9fcb-lgbxz 1/1 Running 0 3h9m
hco-webhook-6c4f88557f-rgqdc 1/1 Running 0 3h9m
hostpath-provisioner-operator-54dcb9dd45-8x5q9 1/1 Running 0 3h3m
hyperconverged-cluster-cli-download-5dfbc88f5-7jbdt 1/1 Running 0 152m
mtq-operator-56d644c994-nrgx4 1/1 Running 0 3h3m
ssp-operator-7d8b9b4c9f-kxs2s 1/1 Running 1 (118m ago) 121m
virt-operator-768f87488d-m45hh 1/1 Running 0 3h3m
virt-operator-768f87488d-wq4z8 1/1 Running 0 3h3m
$ kubectl get csv -n operators
NAME DISPLAY VERSION REPLACES PHASE
kubevirt-hyperconverged-operator.v1.10.1 KubeVirt HyperConverged Cluster Operator 1.10.1 kubevirt-hyperconverged-operator.v1.10.0 SucceededHowever, the ssp-operator deployment still showing error log for no matches for kind "VirtualMachine" in version "kubevirt.io/v1".
Is it because I haven't created CR for the HCO as well as the KubeVirt resource in my cluster?
tiraboschi
commented
2 weeks ago Do you mean the CR for the
HyperConverged, like on this example hco.cr.yaml
Yes, please. The Kubevirt operator will create the VirtualMachine CRD only at that step and that specific version of the SSP operator is going to crash until the the VirtualMachine CRD is there.
kgfathur
commented
2 weeks ago Thanks @tiraboschi for your help!
I try to rebuild my cluster :D then repeat the previous steps. Currently, I don't understand all of the config parameters from sample manifest hco.cr.yaml, so I am creating simple CR for the HCO:
apiVersion: v1
kind: Namespace
metadata:
name: kubevirt-hyperconverged
---
apiVersion: hco.kubevirt.io/v1beta1
kind: HyperConverged
metadata:
name: kubevirt-hyperconverged
namespace: kubevirt-hyperconverged
spec: {}I check the kubevirt, cdi, and networkaddonsconfigs already deployed.
$ kubectl get hco -A
NAMESPACE NAME AGE
kubevirt-hyperconverged kubevirt-hyperconverged 7m51s
$ kubectl get kubevirts.kubevirt.io -A
NAMESPACE NAME AGE PHASE
kubevirt-hyperconverged kubevirt-kubevirt-hyperconverged 7m56s
$ kubectl get cdi -A
NAME AGE PHASE
cdi-kubevirt-hyperconverged 8m2s Deployed
$ kubectl get networkaddonsconfigs.networkaddonsoperator.network.kubevirt.io
NAME AGE
cluster 8m15sIn the operators namespace there are additional pod/deployment.
$ kubectl get pod -n operators
NAME READY STATUS RESTARTS AGE
cdi-apiserver-7c6db74f64-w9zgk 1/1 Running 0 7m41s
cdi-deployment-6b7ff46f6c-sg7pd 1/1 Running 0 7m41s
cdi-operator-7d9ff47dcc-vjk56 1/1 Running 0 12m
cdi-uploadproxy-865dd5c74c-sd7jv 1/1 Running 0 7m40s
cluster-network-addons-operator-5b54878c48-gtkvb 2/2 Running 0 12m
hco-operator-7fd5bd9fcb-zkkl7 1/1 Running 0 12m
hco-webhook-b566bf99d-hth7x 1/1 Running 0 12m
hostpath-provisioner-operator-559ddbf7c4-vrp75 1/1 Running 0 12m
hyperconverged-cluster-cli-download-5dfbc88f5-jpt87 1/1 Running 0 12m
kubemacpool-cert-manager-7c4b6fd4d-lfwvk 1/1 Running 0 7m40s
kubemacpool-mac-controller-manager-6cf8d5564c-5fnnm 2/2 Running 0 7m40s
mtq-operator-56d644c994-wv96p 1/1 Running 0 12m
ssp-operator-857d5dd7d8-8pwj5 0/1 CrashLoopBackOff 4 (9s ago) 12m
virt-operator-768f87488d-lk29r 1/1 Running 0 12m
virt-operator-768f87488d-wg9gd 1/1 Running 0 12mHowever, there's nothing created on kubevirt-hyperconverged namespace:
$ kubectl get all,secret,cm,sa -n kubevirt-hyperconverged
NAME DATA AGE
configmap/kube-root-ca.crt 1 42m
NAME SECRETS AGE
serviceaccount/default 0 42mthe VirtualMachine's CRD also still not exists:
$ kubectl get crd | grep -i virt
cdiconfigs.cdi.kubevirt.io 2024-06-24T16:09:30Z
cdis.cdi.kubevirt.io 2024-06-24T16:04:07Z
dataimportcrons.cdi.kubevirt.io 2024-06-24T16:09:30Z
datasources.cdi.kubevirt.io 2024-06-24T16:09:30Z
datavolumes.cdi.kubevirt.io 2024-06-24T16:09:30Z
hostpathprovisioners.hostpathprovisioner.kubevirt.io 2024-06-24T16:04:07Z
hyperconvergeds.hco.kubevirt.io 2024-06-24T16:04:07Z
kubevirts.kubevirt.io 2024-06-24T16:04:07Z
mtqs.mtq.kubevirt.io 2024-06-24T16:04:07Z
networkaddonsconfigs.networkaddonsoperator.network.kubevirt.io 2024-06-24T16:04:07Z
objecttransfers.cdi.kubevirt.io 2024-06-24T16:09:30Z
ssps.ssp.kubevirt.io 2024-06-24T16:04:07Z
storageprofiles.cdi.kubevirt.io 2024-06-24T16:09:30Z
volumeclonesources.cdi.kubevirt.io 2024-06-24T16:09:30Z
volumeimportsources.cdi.kubevirt.io 2024-06-24T16:09:30Z
volumeuploadsources.cdi.kubevirt.io 2024-06-24T16:09:30Z
$ kubectl api-resources | grep -i kubevirt
cdiconfigs cdi.kubevirt.io/v1beta1 false CDIConfig
cdis cdi,cdis cdi.kubevirt.io/v1beta1 false CDI
dataimportcrons dic,dics cdi.kubevirt.io/v1beta1 true DataImportCron
datasources das cdi.kubevirt.io/v1beta1 true DataSource
datavolumes dv,dvs cdi.kubevirt.io/v1beta1 true DataVolume
objecttransfers ot,ots cdi.kubevirt.io/v1beta1 false ObjectTransfer
storageprofiles cdi.kubevirt.io/v1beta1 false StorageProfile
volumeclonesources cdi.kubevirt.io/v1beta1 true VolumeCloneSource
volumeimportsources cdi.kubevirt.io/v1beta1 true VolumeImportSource
volumeuploadsources cdi.kubevirt.io/v1beta1 true VolumeUploadSource
hyperconvergeds hco,hcos hco.kubevirt.io/v1beta1 true HyperConverged
hostpathprovisioners hostpathprovisioner.kubevirt.io/v1beta1 false HostPathProvisioner
kubevirts kv,kvs kubevirt.io/v1 true KubeVirt
mtqs mtq,mtqs mtq.kubevirt.io/v1alpha1 false MTQ
networkaddonsconfigs networkaddonsoperator.network.kubevirt.io/v1 false NetworkAddonsConfig
ssps ssp.kubevirt.io/v1beta2 true SSP
uploadtokenrequests utr,utrs upload.cdi.kubevirt.io/v1beta1 true UploadTokenRequestAnd as expected, the SSP Operator still CrashLoopBackOff:
{"level":"error","ts":"2024-06-24T16:13:54Z","logger":"controller-runtime.source","msg":"if kind is a CRD, it should be installed before calling Start","kind":"VirtualMachine.kubevirt.io","error":"no matches for kind \"VirtualMachine\" in version \"kubevirt.io/v1\"","stacktrace":"sigs.k8s.io/controller-runtime/pkg/source.(*Kind).Start.func1.1\n\t/workspace/vendor/sigs.k8s.io/controller-runtime/pkg/source/source.go:143\nk8s.io/apimachinery/pkg/util/wait.runConditionWithCrashProtectionWithContext\n\t/workspace/vendor/k8s.io/apimachinery/pkg/util/wait/wait.go:154\nk8s.io/apimachinery/pkg/util/wait.waitForWithContext\n\t/workspace/vendor/k8s.io/apimachinery/pkg/util/wait/wait.go:207\nk8s.io/apimachinery/pkg/util/wait.poll\n\t/workspace/vendor/k8s.io/apimachinery/pkg/util/wait/poll.go:260\nk8s.io/apimachinery/pkg/util/wait.PollImmediateUntilWithContext\n\t/workspace/vendor/k8s.io/apimachinery/pkg/util/wait/poll.go:200\nsigs.k8s.io/controller-runtime/pkg/source.(*Kind).Start.func1\n\t/workspace/vendor/sigs.k8s.io/controller-runtime/pkg/source/source.go:136"}
{"level":"error","ts":"2024-06-24T16:14:04Z","msg":"Could not wait for Cache to sync","controller":"vm-controller","controllerGroup":"kubevirt.io","controllerKind":"VirtualMachine","error":"failed to wait for vm-controller caches to sync: timed out waiting for cache to be synced","stacktrace":"sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Start.func2.1\n\t/workspace/vendor/sigs.k8s.io/controller-runtime/pkg/internal/controller/controller.go:211\nsigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Start.func2\n\t/workspace/vendor/sigs.k8s.io/controller-runtime/pkg/internal/controller/controller.go:216\nsigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Start\n\t/workspace/vendor/sigs.k8s.io/controller-runtime/pkg/internal/controller/controller.go:242\nsigs.k8s.io/controller-runtime/pkg/manager.(*runnableGroup).reconcile.func1\n\t/workspace/vendor/sigs.k8s.io/controller-runtime/pkg/manager/runnable_group.go:219"}
{"level":"info","ts":"2024-06-24T16:14:04Z","msg":"Stopping and waiting for non leader election runnables"}
{"level":"info","ts":"2024-06-24T16:14:04Z","logger":"setup","msg":"shutting down Prometheus metrics server"}
{"level":"info","ts":"2024-06-24T16:14:04Z","msg":"Stopping and waiting for leader election runnables"}
{"level":"info","ts":"2024-06-24T16:14:04Z","msg":"Shutdown signal received, waiting for all workers to finish","controller":"ssp","controllerGroup":"ssp.kubevirt.io","controllerKind":"SSP"}
{"level":"info","ts":"2024-06-24T16:14:04Z","msg":"All workers finished","controller":"ssp","controllerGroup":"ssp.kubevirt.io","controllerKind":"SSP"}
{"level":"info","ts":"2024-06-24T16:14:04Z","msg":"Shutdown signal received, waiting for all workers to finish","controller":"service-controller","controllerGroup":"","controllerKind":"Service"}
{"level":"info","ts":"2024-06-24T16:14:04Z","msg":"All workers finished","controller":"service-controller","controllerGroup":"","controllerKind":"Service"}
{"level":"info","ts":"2024-06-24T16:14:04Z","msg":"Stopping and waiting for caches"}
{"level":"info","ts":"2024-06-24T16:14:04Z","msg":"Stopping and waiting for webhooks"}
{"level":"info","ts":"2024-06-24T16:14:04Z","logger":"controller-runtime.webhook","msg":"shutting down webhook server"}
{"level":"info","ts":"2024-06-24T16:14:04Z","msg":"Wait completed, proceeding to shutdown the manager"}
{"level":"error","ts":"2024-06-24T16:14:04Z","msg":"problem running manager","error":"failed to wait for vm-controller caches to sync: timed out waiting for cache to be synced","stacktrace":"kubevirt.io/ssp-operator/controllers.CreateAndStartReconciler\n\t/workspace/controllers/setup.go:42\nmain.main\n\t/workspace/main.go:273\nruntime.main\n\t/usr/local/go/src/runtime/proc.go:250"}
{"level":"error","ts":"2024-06-24T16:14:04Z","logger":"setup","msg":"unable to create or start controller","controller":"SSP","error":"failed to wait for vm-controller caches to sync: timed out waiting for cache to be synced","stacktrace":"main.main\n\t/workspace/main.go:274\nruntime.main\n\t/usr/local/go/src/runtime/proc.go:250"}I will try to re-create the HCO using the entire sample manifest from hco.cr.yaml
Is there any mandatory configs/params on the spec of the HCO's CR? (maybe my previous HCO CR missing mandatory configs, since it just empty/default spec?); or
Is there any additional step needed?
kgfathur
commented
2 weeks ago Update: Applying the hco.cr.yaml from the main branch is not applicable on my deployment.
$ kubectl create -f https://raw.githubusercontent.com/kubevirt/hyperconverged-cluster-operator/main/deploy/hco.cr.yaml
Error from server (BadRequest): error when creating "manifests/kubevirt/hco-cr-sample.yaml": HyperConverged in version "v1beta1" cannot be handled as a HyperConverged: strict decoding error: unknown field "spec.featureGates.alignCPUs", unknown field "spec.featureGates.autoResourceLimits", unknown field "spec.featureGates.downwardMetrics", unknown field "spec.featureGates.enableApplicationAwareQuota", unknown field "spec.higherWorkloadDensity", unknown field "spec.virtualMachineOptions.disableSerialConsoleLog"I think it because the stable channel still version v1.10.1 (Dec 31, 2023)
Trying with hco.cr.yaml v1.10.1
$ kubectl create -f https://raw.githubusercontent.com/kubevirt/hyperconverged-cluster-operator/v1.10.1/deploy/hco.cr.yaml
hyperconverged.hco.kubevirt.io/kubevirt-hyperconverged created
$ kubectl get csv -n operators
NAME DISPLAY VERSION REPLACES PHASE
kubevirt-hyperconverged-operator.v1.10.1 KubeVirt HyperConverged Cluster Operator 1.10.1 kubevirt-hyperconverged-operator.v1.10.0 Succeeded
$ kubectl get hco -A
NAMESPACE NAME AGE
kubevirt-hyperconverged kubevirt-hyperconverged 2m3s
$ kubectl get cdi -A
NAME AGE PHASE
cdi-kubevirt-hyperconverged 2m8s Deployed
$ kubectl get networkaddonsconfigs -A
NAME AGE
cluster 2m15s
$ kubectl get all,secret,cm,sa -n kubevirt-hyperconverged
NAME DATA AGE
configmap/kube-root-ca.crt 1 12m
NAME SECRETS AGE
serviceaccount/default 0 12mNow, the SSP Operator is not CrashLoopBackOff:
$ kubectl get pod -n operators
NAME READY STATUS RESTARTS AGE
cdi-apiserver-7c6db74f64-p59bm 1/1 Running 0 7m30s
cdi-deployment-6b7ff46f6c-lb4lg 1/1 Running 0 7m30s
cdi-operator-7d9ff47dcc-st428 1/1 Running 0 11m
cdi-uploadproxy-865dd5c74c-9g7rp 1/1 Running 0 7m30s
cluster-network-addons-operator-5b54878c48-8wcsg 2/2 Running 0 11m
hco-operator-7fd5bd9fcb-29m65 1/1 Running 0 11m
hco-webhook-68c8d7b486-pvzjs 1/1 Running 0 11m
hostpath-provisioner-operator-686c9c6954-7tzmv 1/1 Running 0 11m
hyperconverged-cluster-cli-download-5dfbc88f5-qtlgw 1/1 Running 0 11m
kubemacpool-cert-manager-7c4b6fd4d-2svmz 1/1 Running 0 7m30s
kubemacpool-mac-controller-manager-6cf8d5564c-xgmxc 2/2 Running 0 7m30s
mtq-operator-56d644c994-s4r6z 1/1 Running 0 11m
ssp-operator-765d67b668-qtw2m 1/1 Running 0 11m
virt-operator-768f87488d-5n5pg 1/1 Running 0 11m
virt-operator-768f87488d-xr27x 1/1 Running 0 11mHowever, the VirtualMachine's CRD still not exists:
$ kubectl get crd | grep -i virt
cdiconfigs.cdi.kubevirt.io 2024-06-24T17:43:04Z
cdis.cdi.kubevirt.io 2024-06-24T17:39:07Z
dataimportcrons.cdi.kubevirt.io 2024-06-24T17:43:04Z
datasources.cdi.kubevirt.io 2024-06-24T17:43:04Z
datavolumes.cdi.kubevirt.io 2024-06-24T17:43:04Z
hostpathprovisioners.hostpathprovisioner.kubevirt.io 2024-06-24T17:39:08Z
hyperconvergeds.hco.kubevirt.io 2024-06-24T17:39:08Z
kubevirts.kubevirt.io 2024-06-24T17:39:08Z
mtqs.mtq.kubevirt.io 2024-06-24T17:39:08Z
networkaddonsconfigs.networkaddonsoperator.network.kubevirt.io 2024-06-24T17:39:07Z
objecttransfers.cdi.kubevirt.io 2024-06-24T17:43:04Z
ssps.ssp.kubevirt.io 2024-06-24T17:39:08Z
storageprofiles.cdi.kubevirt.io 2024-06-24T17:43:04Z
volumeclonesources.cdi.kubevirt.io 2024-06-24T17:43:04Z
volumeimportsources.cdi.kubevirt.io 2024-06-24T17:43:04Z
volumeuploadsources.cdi.kubevirt.io 2024-06-24T17:43:04Z$ kubectl logs -n operators ssp-operator-765d67b668-qtw2m
...
{"level":"error","ts":"2024-06-24T17:49:46Z","logger":"controller-runtime.source","msg":"if kind is a CRD, it should be installed before calling Start","kind":"VirtualMachine.kubevirt.io","error":"no matches for kind \"VirtualMachine\" in version \"kubevirt.io/v1\"","stacktrace":"sigs.k8s.io/controller-runtime/pkg/source.(*Kind).Start.func1.1\n\t/workspace/vendor/sigs.k8s.io/controller-runtime/pkg/source/source.go:143\nk8s.io/apimachinery/pkg/util/wait.runConditionWithCrashProtectionWithContext\n\t/workspace/vendor/k8s.io/apimachinery/pkg/util/wait/wait.go:154\nk8s.io/apimachinery/pkg/util/wait.waitForWithContext\n\t/workspace/vendor/k8s.io/apimachinery/pkg/util/wait/wait.go:207\nk8s.io/apimachinery/pkg/util/wait.poll\n\t/workspace/vendor/k8s.io/apimachinery/pkg/util/wait/poll.go:260\nk8s.io/apimachinery/pkg/util/wait.PollImmediateUntilWithContext\n\t/workspace/vendor/k8s.io/apimachinery/pkg/util/wait/poll.go:200\nsigs.k8s.io/controller-runtime/pkg/source.(*Kind).Start.func1\n\t/workspace/vendor/sigs.k8s.io/controller-runtime/pkg/source/source.go:136"}
{"level":"error","ts":"2024-06-24T17:49:56Z","logger":"controller-runtime.source","msg":"if kind is a CRD, it should be installed before calling Start","kind":"VirtualMachine.kubevirt.io","error":"no matches for kind \"VirtualMachine\" in version \"kubevirt.io/v1\"","stacktrace":"sigs.k8s.io/controller-runtime/pkg/source.(*Kind).Start.func1.1\n\t/workspace/vendor/sigs.k8s.io/controller-runtime/pkg/source/source.go:143\nk8s.io/apimachinery/pkg/util/wait.runConditionWithCrashProtectionWithContext\n\t/workspace/vendor/k8s.io/apimachinery/pkg/util/wait/wait.go:154\nk8s.io/apimachinery/pkg/util/wait.waitForWithContext\n\t/workspace/vendor/k8s.io/apimachinery/pkg/util/wait/wait.go:207\nk8s.io/apimachinery/pkg/util/wait.poll\n\t/workspace/vendor/k8s.io/apimachinery/pkg/util/wait/poll.go:260\nk8s.io/apimachinery/pkg/util/wait.PollImmediateUntilWithContext\n\t/workspace/vendor/k8s.io/apimachinery/pkg/util/wait/poll.go:200\nsigs.k8s.io/controller-runtime/pkg/source.(*Kind).Start.func1\n\t/workspace/vendor/sigs.k8s.io/controller-runtime/pkg/source/source.go:136"}
nunnatsa
commented
2 weeks ago Oh, 1.10.1? this is solved - kind of. See #2954
Please install version 1.11.0, or if not available, 1.10.7
If this is working for you, please let us know. Then we'll close this issue as duplication of #2954, and keep monitor it there.
We still need to fix the catalog to stop suggesting the broken versions.
kgfathur
commented
2 weeks ago After trying version v1.11.0, the ssp-operator working as expected:
$ kubectl create -n kubevirt-hyperconverged -f https://raw.githubusercontent.com/kubevirt/hyperconverged-cluster-operator/v1.10.1/deploy/hco.cr.yaml
hyperconverged.hco.kubevirt.io/kubevirt-hyperconverged created
$ kubectl get csv -n operators
NAME DISPLAY VERSION REPLACES PHASE
kubevirt-hyperconverged-operator.v1.11.0 KubeVirt HyperConverged Cluster Operator 1.11.0 Succeeded
$ kubectl get kubevirts -n kubevirt-hyperconverged
NAME AGE PHASE
kubevirt-kubevirt-hyperconverged 12m
$ kubectl get cdi -n kubevirt-hyperconverged
NAME AGE PHASE
cdi-kubevirt-hyperconverged 12m Deployed
$ kubectl get networkaddonsconfigs
NAME AGE
cluster 12m
$ kubectl get pod -n operators
NAME READY STATUS RESTARTS AGE
aaq-operator-6976859798-2lvbg 1/1 Running 0 43m
cdi-apiserver-64d85bb898-6qkl8 1/1 Running 0 12m
cdi-deployment-78c94b68dc-kxbjj 1/1 Running 0 12m
cdi-operator-5b67c9967b-6x829 1/1 Running 0 43m
cdi-uploadproxy-7779ddfc6b-qd4xt 1/1 Running 0 12m
cluster-network-addons-operator-6c5496f456-6dtw5 2/2 Running 0 65m
hco-operator-68b595d765-p469d 1/1 Running 0 65m
hco-webhook-6c457857cb-sbhxt 1/1 Running 0 65m
hostpath-provisioner-operator-5bc97b8cf5-lhzs6 1/1 Running 0 43m
hyperconverged-cluster-cli-download-fcc89b7d6-kvvdr 1/1 Running 0 65m
kubemacpool-cert-manager-5d7967c84-9n2vr 1/1 Running 0 12m
kubemacpool-mac-controller-manager-7855d88fc7-nb9l4 2/2 Running 0 12m
mtq-operator-79fd98b4b5-wx789 1/1 Running 0 43m
ssp-operator-5c9f855c8-7z7t8 1/1 Running 0 65m
virt-operator-66d6b95fd9-56djg 1/1 Running 0 43m
virt-operator-66d6b95fd9-cnlc5 1/1 Running 0 43m
$ kubectl get all,secret,svc,cm,sa -n kubevirt-hyperconverged
NAME AGE
hyperconverged.hco.kubevirt.io/kubevirt-hyperconverged 22m
NAME AGE PHASE
kubevirt.kubevirt.io/kubevirt-kubevirt-hyperconverged 22m
NAME DATA AGE
configmap/kube-root-ca.crt 1 23m
NAME SECRETS AGE
serviceaccount/default 0 23mThere's no more error related no matches for kind "VirtualMachine" in version "kubevirt.io/v1" on ssp-operator's logs:
$ kubectl logs -n operators ssp-operator-5c9f855c8-7z7t8
...
{"level":"info","ts":"2024-06-25T06:38:01Z","msg":"Starting service-controller"}
{"level":"info","ts":"2024-06-25T06:38:01Z","msg":"Starting EventSource","controller":"ssp","controllerGroup":"ssp.kubevirt.io","controllerKind":"SSP","source":"kind source: *v1beta2.SSP"}
{"level":"info","ts":"2024-06-25T06:38:01Z","msg":"Starting EventSource","controller":"ssp","controllerGroup":"ssp.kubevirt.io","controllerKind":"SSP","source":"kind source: *v1.RoleBinding"}
{"level":"info","ts":"2024-06-25T06:38:01Z","msg":"Starting EventSource","controller":"ssp","controllerGroup":"ssp.kubevirt.io","controllerKind":"SSP","source":"kind source: *v1.ClusterRole"}
{"level":"info","ts":"2024-06-25T06:38:01Z","msg":"Starting EventSource","controller":"ssp","controllerGroup":"ssp.kubevirt.io","controllerKind":"SSP","source":"kind source: *v1.Role"}
{"level":"info","ts":"2024-06-25T06:38:01Z","msg":"Starting EventSource","controller":"ssp","controllerGroup":"ssp.kubevirt.io","controllerKind":"SSP","source":"kind source: *v1.Namespace"}
{"level":"info","ts":"2024-06-25T06:38:01Z","msg":"Starting EventSource","controller":"ssp","controllerGroup":"ssp.kubevirt.io","controllerKind":"SSP","source":"kind source: *v1.ServiceAccount"}
{"level":"info","ts":"2024-06-25T06:38:01Z","msg":"Starting EventSource","controller":"ssp","controllerGroup":"ssp.kubevirt.io","controllerKind":"SSP","source":"kind source: *v1.ConfigMap"}
{"level":"info","ts":"2024-06-25T06:38:01Z","msg":"Starting Controller","controller":"ssp","controllerGroup":"ssp.kubevirt.io","controllerKind":"SSP"}
{"level":"info","ts":"2024-06-25T06:38:01Z","msg":"service-controller started"}
{"level":"info","ts":"2024-06-25T06:38:01Z","msg":"Starting EventSource","controller":"service-controller","controllerGroup":"","controllerKind":"Service","source":"kind source: *v1.Service"}
{"level":"info","ts":"2024-06-25T06:38:01Z","msg":"Starting Controller","controller":"service-controller","controllerGroup":"","controllerKind":"Service"}
{"level":"info","ts":"2024-06-25T06:38:01Z","msg":"Starting workers","controller":"ssp","controllerGroup":"ssp.kubevirt.io","controllerKind":"SSP","worker count":1}
{"level":"info","ts":"2024-06-25T06:38:01Z","msg":"Starting workers","controller":"service-controller","controllerGroup":"","controllerKind":"Service","worker count":1}
{"level":"info","ts":"2024-06-25T06:38:01Z","logger":"controllers.Resources","msg":"Starting service reconciliation...","request":"operators/ssp-operator-metrics"}However, the CRDs related to VirtualMachine is not created yet.
$ kubectl get crd | grep -i virt
aaqs.aaq.kubevirt.io 2024-06-25T06:21:07Z
cdiconfigs.cdi.kubevirt.io 2024-06-25T07:13:35Z
cdis.cdi.kubevirt.io 2024-06-25T06:21:07Z
dataimportcrons.cdi.kubevirt.io 2024-06-25T07:13:35Z
datasources.cdi.kubevirt.io 2024-06-25T07:13:35Z
datavolumes.cdi.kubevirt.io 2024-06-25T07:13:35Z
hostpathprovisioners.hostpathprovisioner.kubevirt.io 2024-06-25T06:21:07Z
hyperconvergeds.hco.kubevirt.io 2024-06-25T06:21:07Z
kubevirts.kubevirt.io 2024-06-25T06:21:07Z
mtqs.mtq.kubevirt.io 2024-06-25T06:21:07Z
networkaddonsconfigs.networkaddonsoperator.network.kubevirt.io 2024-06-25T06:21:07Z
objecttransfers.cdi.kubevirt.io 2024-06-25T07:13:35Z
ssps.ssp.kubevirt.io 2024-06-25T06:21:07Z
storageprofiles.cdi.kubevirt.io 2024-06-25T07:13:35Z
volumeclonesources.cdi.kubevirt.io 2024-06-25T07:13:35Z
volumeimportsources.cdi.kubevirt.io 2024-06-25T07:13:35Z
volumeuploadsources.cdi.kubevirt.io 2024-06-25T07:13:35ZIs there any additional steps to create the VirtualMachine CRDs?
nunnatsa
commented
2 weeks ago virt-api should deploy the CRD, but I can't see virt pods (other than the virt-operator). Could you please check the virt-controller log? to check if there the virt-api deployment was created?
kgfathur
commented
1 week ago virt-api should deploy the CRD, but I can't see virt pods (other than the virt-operator). Could you please check the virt-controller log? to check if there the virt-api deployment was created?
I cannot find the virt-controller component, maybe it's not created yet. Do you mean the virt-operator? because I see some clue from the log bellow.
I deploy the hco.cr.yaml (v1.11.0) from the default sample. Since it's error when installing it on another namespace except the kubevirt-hyperconverged namespace.
$ kubectl create -f https://raw.githubusercontent.com/kubevirt/hyperconverged-cluster-operator/v1.10.1/deploy/hco.cr.yaml
Error from server (Forbidden): error when creating "manifests/kubevirt/hco.cr.v1.11.0.yaml": admission webhook "validate-hco.kubevirt.io" denied the request: invalid namespace for v1beta1.HyperConverged - please use the kubevirt-hyperconverged namespaceSo, I deploy it on the kubevirt-hyperconverged namespace.
$ kubectl create -n kubevirt-hyperconverged -f https://raw.githubusercontent.com/kubevirt/hyperconverged-cluster-operator/v1.10.1/deploy/hco.cr.yaml
hyperconverged.hco.kubevirt.io/kubevirt-hyperconverged createdHowever, there's no pod/deployment at all in the kubevirt-hyperconverged namespace (maybe it's not created yet).
$ kubectl get all,secret,svc,cm,sa -n kubevirt-hyperconverged
NAME AGE
hyperconverged.hco.kubevirt.io/kubevirt-hyperconverged 45h
NAME AGE PHASE
kubevirt.kubevirt.io/kubevirt-kubevirt-hyperconverged 45h
NAME DATA AGE
configmap/kube-root-ca.crt 1 45h
NAME SECRETS AGE
serviceaccount/default 0 45hI install the HCO operator using the default sample manifests for the deployment via OLM from operatorhub.io.
curl -sfL https://operatorhub.io/install/community-kubevirt-hyperconverged.yaml -o community-kubevirt-hyperconverged.yaml
vim community-kubevirt-hyperconverged.yaml # just edit the name and the channel
kubectl create -f community-kubevirt-hyperconverged.yamlapiVersion: operators.coreos.com/v1alpha1
kind: Subscription
metadata:
name: kubevirt-hyperconverged
namespace: operators
spec:
channel: 1.11.0
name: community-kubevirt-hyperconverged
source: operatorhubio-catalog
sourceNamespace: olmI just edit the channel to 1.11.0 as suggested before.
It seems some of the components are deployed on the operators namespace, because from the sample manifest, the subscription created on namespace: operators.
$ kubectl get pod -n operators
NAME READY STATUS RESTARTS AGE
aaq-operator-6976859798-2lvbg 1/1 Running 0 45h
cdi-apiserver-64d85bb898-6qkl8 1/1 Running 0 45h
cdi-deployment-78c94b68dc-kxbjj 1/1 Running 0 45h
cdi-operator-5b67c9967b-6x829 1/1 Running 0 45h
cdi-uploadproxy-7779ddfc6b-qd4xt 1/1 Running 0 45h
cluster-network-addons-operator-6c5496f456-6dtw5 2/2 Running 0 46h
hco-operator-68b595d765-p469d 1/1 Running 0 46h
hco-webhook-6c457857cb-sbhxt 1/1 Running 0 46h
hostpath-provisioner-operator-5bc97b8cf5-lhzs6 1/1 Running 0 45h
hyperconverged-cluster-cli-download-fcc89b7d6-kvvdr 1/1 Running 0 46h
kubemacpool-cert-manager-5d7967c84-9n2vr 1/1 Running 0 45h
kubemacpool-mac-controller-manager-7855d88fc7-nb9l4 2/2 Running 0 45h
mtq-operator-79fd98b4b5-wx789 1/1 Running 0 45h
ssp-operator-5c9f855c8-7z7t8 1/1 Running 0 46h
virt-operator-66d6b95fd9-56djg 1/1 Running 0 45h
virt-operator-66d6b95fd9-cnlc5 1/1 Running 0 45hThen checking for the virt-operator logs found a clue KubeVirt CR is created in another namespace than the operator, that is not supported.
$ kubectl logs -n operators virt-operator-66d6b95fd9-56djg
...
{"component":"virt-operator","kind":"","level":"info","msg":"Handling KubeVirt resource","name":"kubevirt-kubevirt-hyperconverged","namespace":"kubevirt-hyperconverged","pos":"kubevirt.go:727","timestamp":"2024-06-26T21:21:28.721859Z","uid":"0a563009-2e46-4027-b3f6-46978c6a2832"}
{"component":"virt-operator","level":"error","msg":"Will ignore the install request until the situation is resolved.","pos":"kubevirt.go:912","reason":"KubeVirt CR is created in another namespace than the operator, that is not supported.","timestamp":"2024-06-26T21:21:28.722153Z"}I don't know is the title of the issue still relevan. Should I create new issue or just rename the issue title?
nunnatsa
commented
1 week ago The whole installation - operators and CR, must be on the kubevirt-hyperconverged namespace.
nunnatsa
commented
1 week ago @tiraboschi - the operatorhub installation guide, says to install with
$ kubectl create -f https://operatorhub.io/install/stable/community-kubevirt-hyperconverged.yamlthe content of the file is:
apiVersion: operators.coreos.com/v1alpha1
kind: Subscription
metadata:
name: my-community-kubevirt-hyperconverged
namespace: operators # << ======= WRONG NS
spec:
channel: stable
name: community-kubevirt-hyperconverged
source: operatorhubio-catalog
sourceNamespace: olmDo we have any control over this file?
kgfathur
commented
1 week ago The whole installation - operators and CR, must be on the
kubevirt-hyperconvergednamespace.
thanks @nunnatsa for the information.
I see some docs on this repo and find that it should be there, on the kubevirt-hyperconverged namespace.
The installation guide from operatorhub.io and artifacthub.io show the wrong namespace.
nunnatsa
commented
1 week ago @tiraboschi - I think maybe the issue is here? https://github.com/kubevirt/hyperconverged-cluster-operator/blob/6bc3cbc651fadca3f28ed5ec3aeaa7ddcd33492c/pkg/components/components.go#L863
nunnatsa
commented
1 week ago Trying to fix here: https://github.com/kubevirt/hyperconverged-cluster-operator/pull/3013
kgfathur
commented
1 week ago @tiraboschi - the operatorhub installation guide, says to install with
$ kubectl create -f https://operatorhub.io/install/stable/community-kubevirt-hyperconverged.yamlthe content of the file is:
apiVersion: operators.coreos.com/v1alpha1 kind: Subscription metadata: name: my-community-kubevirt-hyperconverged namespace: operators # << ======= WRONG NS spec: channel: stable name: community-kubevirt-hyperconverged source: operatorhubio-catalog sourceNamespace: olmDo we have any control over this file?
Beside of the wrong namespace, I think the installation guide on the operatorhub/artifacthub is missing the OperatorGroup CR.
From this repo docs Deploy HCO via OLM, it shows we must create a namespace, subscription and an OperatorGroup to deploy HCO via OLM.
However, the installation guide/manifest from operatorhub/artifacthub, missing the namespace and OperatorGroup.
# original guide
# $ kubectl create -f https://operatorhub.io/install/community-kubevirt-hyperconverged.yaml
# But currently we need to change the namespace to `kubevirt-hyperconverged`
$ curl -sfL https://operatorhub.io/install/community-kubevirt-hyperconverged.yaml -o community-kubevirt-hyperconverged.yaml
$ vim community-kubevirt-hyperconverged.yaml # edit the namespace (I also edit the channel to test the v1.11.0)
$ kubectl create -f community-kubevirt-hyperconverged.yaml
Error from server (NotFound): error when creating "community-kubevirt-hyperconverged.yaml": namespaces "kubevirt-hyperconverged" not foundAfter we create namespace and then deploy HCO via OLM:
$ kubectl create namespace kubevirt-hyperconverged
namespace/kubevirt-hyperconverged created
$ kubectl create -f community-kubevirt-hyperconverged.yaml
subscription.operators.coreos.com/kubevirt-hyperconverged created
$ kubectl get subscriptions -n kubevirt-hyperconverged
NAME PACKAGE SOURCE CHANNEL
kubevirt-hyperconverged community-kubevirt-hyperconverged operatorhubio-catalog 1.11.0The operator deployment via OLM will still be failed, because the operatorgroup not exists:
kubectl logs -n olm catalog-operator-6fbb6bd9bb-qfrh7
...
E0627 05:57:07.871296 1 queueinformer_operator.go:319] sync "kubevirt-hyperconverged" failed: found 0 operatorGroups, expected 1
time="2024-06-27T05:57:07Z" level=info msg="resolving sources" id=lRzgG namespace=kubevirt-hyperconverged
time="2024-06-27T05:57:07Z" level=info msg="checking if subscriptions need update" id=lRzgG namespace=kubevirt-hyperconverged
time="2024-06-27T05:57:08Z" level=info msg="resolving sources" id=e2O2j namespace=kubevirt-hyperconverged
time="2024-06-27T05:57:08Z" level=info msg="checking if subscriptions need update" id=e2O2j namespace=kubevirt-hyperconverged
E0627 05:57:08.271025 1 queueinformer_operator.go:319] sync "kubevirt-hyperconverged" failed: found 0 operatorGroups, expected 1So, we need to create the namespace and OperatorGroup as described on Deploy HCO via OLM.
If the installation guide or the sample manifest on operatorhub/artifacthub can be improved (to include the namespace & operatorgroup), it would be great and help the new user (like me) that trying to deploy the kubevirt using hyperconverged-cluster-operator.
Thank you!
kgfathur
commented
1 week ago After adjusting manifest for the required resources:
# community-kubevirt-hyperconverged.yaml
apiVersion: v1
kind: Namespace
metadata:
name: kubevirt-hyperconverged
---
apiVersion: operators.coreos.com/v1
kind: OperatorGroup
metadata:
name: kubevirt-hyperconverged-group
namespace: kubevirt-hyperconverged
---
apiVersion: operators.coreos.com/v1alpha1
kind: Subscription
metadata:
name: kubevirt-hyperconverged
namespace: kubevirt-hyperconverged
spec:
channel: 1.11.0 # current stable channel has been updated to v1.11.1, but i want to test the v1.11.0
name: community-kubevirt-hyperconverged
source: operatorhubio-catalog
sourceNamespace: olmAfter creating the required resources, HCO v1.11.0 via OLM successfully deployed:
$ kubectl create -f community-kubevirt-hyperconverged.yaml
namespace/kubevirt-hyperconverged created
operatorgroup.operators.coreos.com/kubevirt-hyperconverged-group created
subscription.operators.coreos.com/kubevirt-hyperconverged created
$ kubectl get subscriptions -n kubevirt-hyperconverged
NAME PACKAGE SOURCE CHANNEL
kubevirt-hyperconverged community-kubevirt-hyperconverged operatorhubio-catalog 1.11.0
$ kubectl get csv -n kubevirt-hyperconverged
NAME DISPLAY VERSION REPLACES PHASE
kubevirt-hyperconverged-operator.v1.11.0 KubeVirt HyperConverged Cluster Operator 1.11.0 SucceededAll of the resources deployed successfully on the kubevirt-hyperconverged as expected.
$ kubectl get pod -n kubevirt-hyperconverged
NAME READY STATUS RESTARTS AGE
aaq-operator-6f78c54bcb-vvr9r 1/1 Running 0 3h9m
bridge-marker-5dm6w 1/1 Running 0 170m
bridge-marker-6rwkf 1/1 Running 0 170m
bridge-marker-77ttb 1/1 Running 0 170m
bridge-marker-92fkh 1/1 Running 0 170m
bridge-marker-98zz7 1/1 Running 0 170m
bridge-marker-btpds 1/1 Running 0 170m
bridge-marker-cwt7p 1/1 Running 0 170m
bridge-marker-l2br4 1/1 Running 0 170m
bridge-marker-nqktw 1/1 Running 0 170m
bridge-marker-ps827 1/1 Running 0 170m
bridge-marker-rstl9 1/1 Running 0 170m
bridge-marker-wcqgn 1/1 Running 0 170m
cdi-apiserver-64d85bb898-kmnrn 1/1 Running 0 170m
cdi-deployment-78c94b68dc-tdlmr 1/1 Running 0 170m
cdi-operator-9dc84cdfd-967gl 1/1 Running 0 3h9m
cdi-uploadproxy-7779ddfc6b-vn4kt 1/1 Running 0 170m
cluster-network-addons-operator-59dddbff8d-p7sx8 2/2 Running 0 3h31m
hco-operator-67fdbfdd47-tfjp4 1/1 Running 0 3h31m
hco-webhook-b76849c99-g84st 1/1 Running 0 3h31m
hostpath-provisioner-operator-9f9cdc4c4-m2h24 1/1 Running 0 3h9m
hyperconverged-cluster-cli-download-78c79cd978-bwl65 1/1 Running 0 3h31m
kube-cni-linux-bridge-plugin-2lknr 1/1 Running 0 170m
kube-cni-linux-bridge-plugin-48bx6 1/1 Running 0 170m
kube-cni-linux-bridge-plugin-7qqdb 1/1 Running 0 170m
kube-cni-linux-bridge-plugin-ckqk2 1/1 Running 0 170m
kube-cni-linux-bridge-plugin-j6vlt 1/1 Running 0 170m
kube-cni-linux-bridge-plugin-krx9n 1/1 Running 0 170m
kube-cni-linux-bridge-plugin-l9vf9 1/1 Running 0 170m
kube-cni-linux-bridge-plugin-mrlbq 1/1 Running 0 170m
kube-cni-linux-bridge-plugin-rd297 1/1 Running 0 170m
kube-cni-linux-bridge-plugin-rxszf 1/1 Running 0 170m
kube-cni-linux-bridge-plugin-txtx9 1/1 Running 0 170m
kube-cni-linux-bridge-plugin-zhbr5 1/1 Running 0 170m
kubemacpool-cert-manager-5d7967c84-x5sdw 1/1 Running 0 170m
kubemacpool-mac-controller-manager-7855d88fc7-22kbz 2/2 Running 0 170m
mtq-operator-9b674769-qmksv 1/1 Running 0 3h9m
multus-2nfmr 1/1 Running 0 170m
multus-5kghg 1/1 Running 0 170m
multus-9pr8f 1/1 Running 0 170m
multus-bsz6c 1/1 Running 0 170m
multus-f4kds 1/1 Running 0 170m
multus-fksfd 1/1 Running 0 170m
multus-gdwtj 1/1 Running 0 170m
multus-kp5s5 1/1 Running 0 170m
multus-n795x 1/1 Running 0 170m
multus-p7xbp 1/1 Running 0 170m
multus-t98l5 1/1 Running 0 170m
multus-vt5fs 1/1 Running 0 170m
ssp-operator-7486bd78fc-xbnz9 1/1 Running 1 (169m ago) 3h31m
virt-api-56bdddd94-76bhr 1/1 Running 0 169m
virt-api-56bdddd94-9r67q 1/1 Running 0 169m
virt-controller-5956594b98-547cf 1/1 Running 0 165m
virt-controller-5956594b98-zthrz 1/1 Running 0 165m
virt-exportproxy-57968cd7bc-l6tgb 1/1 Running 0 165m
virt-exportproxy-57968cd7bc-vlndt 1/1 Running 0 165m
virt-handler-9ddjs 1/1 Running 0 165m
virt-handler-kn6g6 1/1 Running 0 165m
virt-handler-kxdhz 1/1 Running 0 165m
virt-handler-lkzgp 1/1 Running 0 165m
virt-handler-rj7md 1/1 Running 0 165m
virt-handler-w95d4 1/1 Running 0 165m
virt-operator-6f6c884b65-pd7w4 1/1 Running 0 3h9m
virt-operator-6f6c884b65-qgg86 1/1 Running 0 3h9mSuccessfully test to create DataVolume and VirtualMachine:
$ kubectl create -f vm.fedora.cdi.yaml
datavolume.cdi.kubevirt.io/fedora-cdi created
virtualmachine.kubevirt.io/fedora-cdi created
$ kubectl get dv
NAME PHASE PROGRESS RESTARTS AGE
fedora-cdi Succeeded 100.0% 1 2m24s
$ kubectl get vm
NAME AGE STATUS READY
fedora-cdi 2m31s Running True
$ virtctl console fedora-cdi
Successfully connected to fedora-cdi console. The escape sequence is ^]
fedora-cdi login: root
Password:
[root@fedora-cdi ~]# uname -a
Linux fedora-cdi 6.8.5-301.fc40.x86_64 #1 SMP PREEMPT_DYNAMIC Thu Apr 11 20:00:10 UTC 2024 x86_64 GNU/Linux
[root@fedora-cdi ~]#Thank you @tiraboschi @nunnatsa for your help! Now, I see the stable channel on operatorhub has been fixed & updated to v1.11.1 as #3009, will test it also.
tiraboschi
commented
1 week ago Do we have any control over this file?
Beside of the wrong namespace, I think the installation guide on the operatorhub/artifacthub is missing the
OperatorGroupCR. From this repo docs Deploy HCO via OLM, it shows we must create a namespace, subscription and an OperatorGroup to deploy HCO via OLM.However, the installation guide/manifest from operatorhub/artifacthub, missing the namespace and
OperatorGroup.# original guide # $ kubectl create -f https://operatorhub.io/install/community-kubevirt-hyperconverged.yaml # But currently we need to change the namespace to `kubevirt-hyperconverged` $ curl -sfL https://operatorhub.io/install/community-kubevirt-hyperconverged.yaml -o community-kubevirt-hyperconverged.yaml $ vim community-kubevirt-hyperconverged.yaml # edit the namespace (I also edit the channel to test the v1.11.0) $ kubectl create -f community-kubevirt-hyperconverged.yaml Error from server (NotFound): error when creating "community-kubevirt-hyperconverged.yaml": namespaces "kubevirt-hyperconverged" not found
That text is somehow hard-coded into the frontend modal of the operatorhub website, see: https://github.com/operator-framework/operatorhub.io/blob/fcee0e77a41f01775f62bd9a840f0787c3e7a951/frontend/src/components/modals/InstallModal.tsx#L105
And since we support also the AllNamespaces (for a different reason) the operatorhub website is considering us as a globalOperator:
https://github.com/operator-framework/operatorhub.io/blob/fcee0e77a41f01775f62bd9a840f0787c3e7a951/frontend/src/utils/operatorUtils.ts#L93
so it's suggesting to install it in the operators namespace where a generic OperatorGroup is already available and so that step is also skipped.
We need to understand if and how we can publish a custom install help in the OperatorHub website.
Thank you @tiraboschi @nunnatsa for your help! Now, I see the stable channel on operatorhub has been fixed & updated to v1.11.1 as https://github.com/kubevirt/hyperconverged-cluster-operator/pull/3009, will test it also.
Thanks for having reported it, I'm going to close this.
What happened: I am deploying HCO v1.10.1 via OLM: https://operatorhub.io/operator/community-kubevirt-hyperconverged. HCO Deployment via OLM on kubernetes (k3s) is failing due to SSP Operator CrashLoopBackOff with error message: no matches for kind "VirtualMachine" in version "kubevirt.io/v1"
I see some other related issues:
2129
https://github.com/kubevirt/ssp-operator/pull/437
What you expected to happen: Successfully deploy HCO via OLM on plain kubernetes (k3s)
How to reproduce it (as minimally and precisely as possible): Install the OLM:
Deploy
kubevirt-hyperconverged:file
community-kubevirt-hyperconverged.yaml(just edit the name):Additional context: I have successfully deployed and testing manual deployment of KubeVirt & CDI Operator (Not HCO) on the same kubernetes k3s cluster & architecture. The Kubernetes k3s cluster already re-deployed (to make sure it's clean without previous kubevirt resources/crds)
Environment:
virtctl version): N/A (kubevirt-hyperconverged-operator v1.10.1)kubectl version):v1.28.9+k3s1KVM with Nested VirtualizationUbuntu 22.04.4 LTS (Jammy Jellyfish)uname -a):Linux 5.15.0-112-generic #122-Ubuntu SMP Thu May 23 07:48:21 UTC 2024 x86_64 x86_64 x86_64 GNU/LinuxCalico with Multus, Ingress:nginx, Storage:ceph-csi-rbd&ceph-csi-cephfs