You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm using mpi-operator v2 in my project, and I also need training-operator TF and Pytorch jobs.
After upgrading training-operator to v1.5, and updating the kube-openapi module to a working reference,
now the mpi-operator returns this errors: ../../go/pkg/mod/github.com/kubeflow/mpi-operator/v2@v2.0.0-20220831182455-5be2a42bf5af/pkg/apis/kubeflow/v2beta1/openapi_generated.go:44:11: cannot use spec.Schema{…} (value of type "github.com/go-openapi/spec".Schema) as type "k8s.io/kube-openapi/pkg/validation/spec".Schema in struct literal
At first I had this line: k8s.io/kube-openapi => k8s.io/kube-openapi v0.0.0-20210113233702-8566a335510f
After the update k8s.io/kube-openapi => k8s.io/kube-openapi v0.0.0-20220124234850-424119656bbf
It looks like an issue or either one of them. Either use the mpi v2 or the tf 1.5.
We don't want to upgrade to mpi-v1 due to workloads in scale.
Are you aware of it? Is there anything that can be done?
The text was updated successfully, but these errors were encountered:
Hi,
I'm using mpi-operator v2 in my project, and I also need training-operator TF and Pytorch jobs.
After upgrading training-operator to v1.5, and updating the kube-openapi module to a working reference,
now the mpi-operator returns this errors:
../../go/pkg/mod/github.com/kubeflow/mpi-operator/v2@v2.0.0-20220831182455-5be2a42bf5af/pkg/apis/kubeflow/v2beta1/openapi_generated.go:44:11: cannot use spec.Schema{…} (value of type "github.com/go-openapi/spec".Schema) as type "k8s.io/kube-openapi/pkg/validation/spec".Schema in struct literal
At first I had this line:
k8s.io/kube-openapi => k8s.io/kube-openapi v0.0.0-20210113233702-8566a335510f
After the update
k8s.io/kube-openapi => k8s.io/kube-openapi v0.0.0-20220124234850-424119656bbf
It looks like an issue or either one of them. Either use the mpi v2 or the tf 1.5.
We don't want to upgrade to mpi-v1 due to workloads in scale.
Are you aware of it? Is there anything that can be done?
The text was updated successfully, but these errors were encountered: