- Clone the Pravega GRPC Gateway repository.
cd
git clone https://github.com/pravega/pravega-grpc-gateway
cd pravega-grpc-gateway
-
Edit the file charts/pravega-grpc-gateway/values.yaml as needed. You will need to set
pravega.controller
andservice.annotations.external-dns.alpha.kubernetes.io/hostname
to match your environment. -
Build and then deploy using Helm.
export DOCKER_REPOSITORY=claudiofahey
export IMAGE_TAG=0.7.0
export NAMESPACE=examples
scripts/build-k8s-components.sh
scripts/deploy-k8s-components.sh
-
Copy secret-example.yaml to secret.yaml.
-
Use
openssl rand -hex 32
to generate two secrets and place them in secret.yaml. -
Edit the file config.yaml. You will need to set
proxy.service.annotations.external-dns.alpha.kubernetes.io/hostname
to match your environment. -
Deploy using Helm.
export NAMESPACE=examples
./deploy-k8s-components.sh
- Open your browser to the proxy service address you set above (http). Alternatively, you may determine the external IP address with the following command.
kubectl get svc/proxy-public -n examples
-
Login with any user name ("videodemo" is recommended). The password is in the file secret.yaml, field
auth.dummy.password
. -
Choose Jupyter Tensorflow Notebook.
-
A session will be created for each user name.
- In Jupyter, open a terminal.
cd data-project
git clone https://github.com/pravega/video-samples
-
Open the notebook
data-project/video-samples/jupyterhub/notebooks/install_dependencies.ipynb
. Click Kernel -> Restart Kernel and Run All Cells. -
Open the notebook
video_data_generator.ipynb
. Click Kernel -> Restart Kernel and Run All Cells. -
Repeat for
video_player.ipynb
andvideo_player_index.ipynb
.