The Semantic Hub is a logical and architectural component of Tractus-X. The source code under this folder contains reference implementations of the SLDT Semantic Hub
Run mvn install
to run unit tests, build and install the package.
To check whether the build was successful, you can start the resulting JAR file from the build process by running java -jar target/semantic-hub-backend-{current-version}.jar
.
Run docker build -t semantic-hub .
In case you want to publish your image into a remote container registry, apply the tag accordingly and docker push
the image.
If you have a running Kubernetes cluster available, you can deploy the Semantic Hub using our Helm Chart, which is located under ./charts/semantic-hub
.
In case you don't have a running cluster, you can set up one by yourself locally, using minikube.
In the following, we will use a minikube cluster for reference.
Before deploying the Semantic Hub, enable a few add-ons in your minikube cluster by running the following commands:
minikube addons enable storage-provisioner
minikube addons enable default-storageclass
minikube addons enable ingress
If you want to use the in-memory triple store that is not persistent (useful for local deployments) set embeddedTripleStore: true
.
In order to deploy the helm chart, first create a new namespace "semantics": kubectl create namespace semantics
.
Then run helm install hub -n semantics ./charts/semantic-hub
. This will set up a new helm deployment in the semantics namespace. By default, the deployment contains the Semantic Hub instance itself, and a Fuseki Triplestore.
Check that the two containers are running by calling kubectl get pod -n semantics
.
To access the Semantic Hub API from the host, you need to configure the Ingress
resource.
By default, the Semantic Hub includes an Ingress
that exposes the API on https://minikube/semantics/hub
For that to work, you need to append /etc/hosts
by running echo "$(minikube ip) minikube" | sudo tee -a /etc/hosts
.
For automated certificate generation, use and configure cert-manager.
By default, authentication is deactivated, please adjust hub.authentication
if needed
The Helm Chart can be configured using the following parameters (incomplete list). For a full overview, please see the values.yaml.
Parameter | Description | Default value |
---|---|---|
hub.image |
The registry and image of the Semantic Hub | semantic-hub:latest |
hub.embeddedTripleStore |
Configures whether the Semantic Hub uses its non-persistent embedded in-memory Triplestore. If set to true , no separate Fuseki instance is deployed. |
false |
hub.host |
This value is used by the Ingress object (if enabled) to route traffic. |
minikube |
hub.authentication |
Enables OAuth2 based authentication/authorization. | false |
hub.idpIssuerUri |
The issuer URI of the OAuth2 identity provider. | http://localhost:8080/auth/realms/catenax |
hub.graphdbBaseUrl |
The URL of an external GraphDB. Ignored if graphdb.enabled is set to false |
http://graphdb:3030 |
hub.ingress.enabled |
Configures if an Ingress resource is created. |
true |
hub.ingress.tls |
Configures whether the Ingress should include TLS configuration. In that case, a separate Secret (as defined by hub.ingress.tlsSecretName ) needs to be provided manually or by using cert-manager |
true |
hub.ingress.tlsSecretName |
The Secret name that contains a tls.crt and tls.key entry. Subject Alternative Name must match the hub.host |
hub-certificate-secret |
hub.ingress.urlPrefix |
The url prefix that is used by the Ingress resource to route traffic |
/semantics/hub |
hub.ingress.className |
The Ingress class name |
nginx |
hub.ingress.annotations |
Annotations to further configure the Ingress resource, e.g. for using with cert-manager . |
Parameter | Description | Default value |
---|---|---|
graphdb.enabled |
Configures, whether a separate Fuseki Triplestore should be deployed in the cluster. | true |
graphdb.storageClassName |
Defines the storage class name of the PersistentVolumeClaim that is used to persist the GraphDB data. |
standard |
graphdb.storageSize |
Size of the PersistentVolumeClaim |
50Gi |