Deployment of Kubeflow fails after running make apply from kubeflow folder with the error message:
...
serviceaccount/kserve-models-web-app created
unable to recognize "build/serving.kserve.io_v1alpha1_clusterservingruntime_kserve-lgbserver.yaml": no matches for kind "ClusterServingRuntime" in version "serving.kserve.io/v1alpha1"
unable to recognize "build/serving.kserve.io_v1alpha1_clusterservingruntime_kserve-mlserver.yaml": no matches for kind "ClusterServingRuntime" in version "serving.kserve.io/v1alpha1"
unable to recognize "build/serving.kserve.io_v1alpha1_clusterservingruntime_kserve-paddleserver.yaml": no matches for kind "ClusterServingRuntime" in version "serving.kserve.io/v1alpha1"
unable to recognize "build/serving.kserve.io_v1alpha1_clusterservingruntime_kserve-pmmlserver.yaml": no matches for kind "ClusterServingRuntime" in version "serving.kserve.io/v1alpha1"
unable to recognize "build/serving.kserve.io_v1alpha1_clusterservingruntime_kserve-sklearnserver.yaml": no matches for kind "ClusterServingRuntime" in version "serving.kserve.io/v1alpha1"
unable to recognize "build/serving.kserve.io_v1alpha1_clusterservingruntime_kserve-tensorflow-serving.yaml": no matches for kind "ClusterServingRuntime" in version "serving.kserve.io/v1alpha1"
unable to recognize "build/serving.kserve.io_v1alpha1_clusterservingruntime_kserve-torchserve.yaml": no matches for kind "ClusterServingRuntime" in version "serving.kserve.io/v1alpha1"
unable to recognize "build/serving.kserve.io_v1alpha1_clusterservingruntime_kserve-tritonserver.yaml": no matches for kind "ClusterServingRuntime" in version "serving.kserve.io/v1alpha1"
unable to recognize "build/serving.kserve.io_v1alpha1_clusterservingruntime_kserve-xgbserver.yaml": no matches for kind "ClusterServingRuntime" in version "serving.kserve.io/v1alpha1"
make[1]: *** [Makefile:13: apply] Error 1
make[1]: Leaving directory '.../kubeflow/contrib/kserve'
make: *** [Makefile:83: apply] Error 1
However, after rerunning make apply the problem resolves.
Should we consider moving KServe CRDs and applying them separately before deploying the runtime resources to have a gap in time between.
gkcalat
commented
2 years ago
Deployment of Kubeflow fails after running
make applyfrom kubeflow folder with the error message:However, after rerunning
make applythe problem resolves.Should we consider moving KServe CRDs and applying them separately before deploying the runtime resources to have a gap in time between.