studyJob controller is blocked by bad CR manifests

[hougangliu]

refer to same root cause . we can follow the recommendation in this comment

ERROR: logging before flag.Parse: E0103 11:58:27.704970       1 streamwatcher.go:109] Unable to decode an event from the watch stream: unable to decode watch event: v1alpha1.StudyJob.Spec: v1alpha1.StudyJobSpec.RequestCount: readUint64: unexpected character: �, error found in #10 byte of ...|stcount":"l","studyN|..., bigger context ...|","parametertype":"categorical"}],"requestcount":"l","studyName":"random-example","suggestionSpec":{|...
ERROR: logging before flag.Parse: E0103 11:58:27.713261       1 streamwatcher.go:109] Unable to decode an event from the watch stream: unable to decode watch event: v1alpha1.StudyJob.Spec: v1alpha1.StudyJobSpec.RequestCount: readUint64: unexpected character: �, error found in #10 byte of ...|stcount":"l","studyN|..., bigger context ...|","parametertype":"categorical"}],"requestcount":"l","studyName":"random-example","suggestionSpec":{|...
ERROR: logging before flag.Parse: W0103 11:58:27.713500       1 reflector.go:341] github.com/kubeflow/katib/vendor/sigs.k8s.io/controller-runtime/pkg/cache/internal/informers_map.go:106: watch of *v1alpha1.StudyJob ended with: very short watch: github.com/kubeflow/katib/vendor/sigs.k8s.io/controller-runtime/pkg/cache/internal/informers_map.go:106: Unexpected watch close - watch lasted less than a second and no items received
ERROR: logging before flag.Parse: E0103 11:58:28.716347       1 reflector.go:205] github.com/kubeflow/katib/vendor/sigs.k8s.io/controller-runtime/pkg/cache/internal/informers_map.go:106: Failed to list *v1alpha1.StudyJob: v1alpha1.StudyJobList.Items: []v1alpha1.StudyJob: v1alpha1.StudyJob.Spec: v1alpha1.StudyJobSpec.RequestCount: readUint64: unexpected character: �, error found in #10 byte of ...|stcount":"l","studyN|..., bigger context ...|","parametertype":"categorical"}],"requestcount":"l","studyName":"random-example","suggestionSpec":{|...
ERROR: logging before flag.Parse: E0103 11:58:29.718285       1 reflector.go:205] github.com/kubeflow/katib/vendor/sigs.k8s.io/controller-runtime/pkg/cache/internal/informers_map.go:106: Failed to list *v1alpha1.StudyJob: v1alpha1.StudyJobList.Items: []v1alpha1.StudyJob: v1alpha1.StudyJob.Spec: v1alpha1.StudyJobSpec.RequestCount: readUint64: unexpected character: �, error found in #10 byte of ...|stcount":"l","studyN|..., bigger context ...|","parametertype":"categorical"}],"requestcount":"l","studyName":"random-example","suggestionSpec":{|...

[johnugeorge]

Job Operators have solved it using unstructured implementation

[hougangliu]

Job Operators have solved it using unstructured implementation

@johnugeorge do you mean this PR for tfjob?
Katib should follow this PR to fix this issue.

[johnugeorge]

Yes. However, katib uses controller-runtime for controller implementation. Need to see how can it be adapted.

@gaocegege Thoughts?

[hougangliu]

/assign