A Pipeline is set of Tasks. Tasks run in parallel. The execution flow can be controlled implicitly (via one task consume a result of another) or explcitly with mechanisms like runAfter, when and finally. A Task is a sequence of Steps. Steps run sequentially. The step can programmatically determine to execute or skip.

To execute a Pipeline create a PipelineRun, an object that identifies:

• the Pipeline to execute and
• the values of any parameters

Tekton creates a TaskRun for each Task in the Pipeline. A TaskRun is an object that identifies:

• the Task and
• the values of any parameters (passed from the PipelineRun)

The TaskRun is implemented by a Pod Each Step is implemented by a Container in the Pod.

1. HF token
2. s3 bucket and necessary keys for uploading results
3. Access to cluster with tekton (installing Tekton)

1. Create a namespace where the Tekton pipeline will execute.

   export $NAMESPACE=your_namespace

   kubectl create ns $NAMESPACE

   or

   oc new-project $NAMESPACE

   For convenience, set the current context:

   kubectl config set-context --current --namespace $NAMESPACE

2. Create a secret hf-secret containing your HuggingFace token in the namespace.

   kubectl create secret generic hf-secret \
       --namespace ${NAMESPACE} \
       --from-literal="HF_TOKEN=${HF_TOKEN}" \
       --dry-run=client -o yaml | kubectl apply -f -

3. Create a secret containing your s3 credentials AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY.

   kubectl create secret generic s3-secret \
       --namespace ${NAMESPACE} \
       --from-literal="AWS_ACCESS_KEY_ID=${AWS_ACCESS_KEY_ID}" \
       --from-literal="AWS_SECRET_ACCESS_KEY=${AWS_SECRET_ACCESS_KEY}" \
       --dry-run=client -o yaml | kubectl apply -f -

4. Give the tasks needed permissions

   envsubst '$NAMESPACE' < tekton/roles.yaml | kubectl apply -f -

   oc adm policy add-scc-to-user anyuid -z default -n $NAMESPACE

5. Create RWX PVC model-pvc (300Gi) and data-pvc (20Gi) and source-pvc (20Gi) for storing models and execution results, respectively. These PVC is shared between all tasks. For example:

   export PVC_NAME=model-pvc
   export PVC_SIZE=300Gi

   export PVC_NAME=data-pvc
   export PVC_SIZE=20Gi

   export PVC_NAME=source-pvc
   export PVC_SIZE=20Gi

   cat <<EOF | kubectl apply -f -
   apiVersion: v1
   kind: PersistentVolumeClaim
   metadata:
       name: ${PVC_NAME}
       namespace: ${NAMESPACE}
   spec:
       accessModes:
       - ReadWriteMany
       resources:
           requests:
               storage: ${PVC_SIZE}
       # storageClassName: ocs-storagecluster-cephfs
       volumeMode: Filesystem
   EOF

6. Install tkn cli:

   brew install tektoncd-cli

1. Deploy the steps and tasks:

   for step in tekton/steps/*.yaml; do
       kubectl apply -f $step
   done
   for task in tekton/tasks/*.yaml; do
       kubectl apply -f $task
   done

2. Build and deploy the pipeline:

   python tektonc/tektonc.py \
   -t tektoncsample/prefix-caching/pipeline.yaml.j2 \
   -f tektoncsample/prefix-caching/values.yaml \
   -r tektoncsample/prefix-caching/pipelinerun.yaml \
   -o tektoncsample/prefix-caching/pipeline.yaml
   
   kubectl apply -f tektoncsample/prefix-caching/pipeline.yaml

3. Deploy the PipelineRun.

   Run the pipeline by deploying the PipelineRun:

   kubectl apply -f tektoncsample/prefix-caching/pipelinerun.yaml

   export EXPERIMENT_ID=
   export NAMESPACE=
   envsubst < tektoncsample/prefix-caching/pipelinerun.yaml | kubectl apply -f -

See the PipelineRun object created:

tkn pr list

See the TaskRun objects created:

tkn tr list

See the logs for a TaskRun:

tkn tr logs <taskrun_name> -f

Describe a TaskRun:

tkn tr describe <taskrun_name>

Delete the PipelineRun:

tkn pr delete <pipelinerun_name> -f