ODH Operator is introducing new CRD called DataScienceCluster. The new feature set will be
released in phases and will be made available before release in the form of a custom
Operator catalog
- ODH Operator team will provide new catalogsource image with tag corresponding to latest
pre-release
in ODH releases.
Alternatively, you can directly get the preview version
export RELEASE_TAG=$( curl https://api.github.com/repos/opendatahub-io/opendatahub-operator/releases | jq -r 'map(select(.prerelease)) | first | .tag_name')
- Deploy CatalogSource
cat <<EOF | oc create -f -
apiVersion: operators.coreos.com/v1alpha1
kind: CatalogSource
metadata:
name: opendatahub-dev-catalog
namespace: openshift-marketplace
spec:
sourceType: grpc
image: quay.io/opendatahub/opendatahub-operator-catalog:$RELEASE_TAG
displayName: Open Data Hub Operator (Preview)
publisher: ODH
EOF
- Subscribe to the ODH custom catalog
cat <<EOF | oc create -f -
apiVersion: operators.coreos.com/v1alpha1
kind: Subscription
metadata:
name: opendatahub-operator
namespace: openshift-operators
spec:
channel: fast
name: opendatahub-operator
source: opendatahub-dev-catalog
sourceNamespace: openshift-marketplace
EOF
- Apply updated version of catalog
export NEW_RELEASE_TAG=$( curl https://api.github.com/repos/opendatahub-io/opendatahub-operator/releases | jq -r 'map(select(.prerelease)) | first | .tag_name')
cat <<EOF | oc apply -f -
apiVersion: operators.coreos.com/v1alpha1
kind: CatalogSource
metadata:
name: opendatahub-dev-catalog
namespace: openshift-marketplace
spec:
sourceType: grpc
image: quay.io/opendatahub/opendatahub-operator-catalog:$NEW_RELEASE_TAG
displayName: Open Data Hub Operator (Preview)
publisher: ODH
EOF
- Select
fast
channel to update
- When Operator is installed it creates a namespace called
opendatahub
. - Users need to create required
DataScienceCluster
resource by going to theInstalled Operators
tab in the OpenShift Cluster. - Users should explicitly set components with
enabled: true
in order for components to be installed.
cat <<EOF | oc apply -f -
apiVersion: datasciencecluster.opendatahub.io/v1alpha1
kind: DataScienceCluster
metadata:
name: example
spec:
components:
codeflare:
enabled: true
dashboard:
enabled: true
datasciencepipelines:
enabled: true
kserve:
enabled: true
modelmeshserving:
enabled: true
ray:
enabled: true
workbenches:
enabled: true
EOF
- Currently on integration of ODH core components is available with the Operator.
- Tier 1 and Tier 2 components can be deployed manually using kustomize build