Deploy with Helm#

Prerequisites#

System Requirements
K8s installation on single or multi node must be done as pre-requisite to continue the following deployment. Note: The kubernetes cluster is set up with kubeadm, kubectl and kubelet packages on single and multi nodes with v1.30.2. Refer to tutorials such as https://dev.to/korakrit/installing-kubernetes-single-node-setup-on-ubuntu-2404-4f47 and many other online tutorials to set up Kubernetes cluster on the web with host OS as Ubuntu 24.04.
For Helm installation, refer to Helm website

Note If Ubuntu Desktop is not installed on the target system, follow the instructions from Ubuntu to install Ubuntu desktop.

Access to the helm charts - use one of the below options#

Use helm charts available at edge-ai-libraries/microservices/time-series-analytics/helm
Using pre-built helm charts:

Follow this procedure on the target system to install the package.
1. Download the Helm chart with the following command
  
  helm pull oci://registry-1.docker.io/intel/ia-time-series-analytics-microservice --version 2026.1.0-rc1-helm
2. Extract the package using the following command
  
  tar -xvzf ia-time-series-analytics-microservice-2026.1.0-rc1-helm.tgz
- Get into the Helm directory
  
  cd ia-time-series-analytics-microservice

Install Helm Charts#

Note:

Uninstall the Helm charts if already installed.

If the worker nodes are running behind proxy server, then please additionally set env.HTTP_PROXY and env.HTTPS_PROXY env like the way env.TELEGRAF_INPUT_PLUGIN is being set as follows with helm install command

cd edge-ai-libraries/microservices/time-series-analytics/helm # path relative to git clone folder
# Install helm charts
helm install time-series-analytics-microservice . -n apps --create-namespace

Use the following command to verify if all the application resources got installed w/ their status:

   kubectl get all -n apps

Upload the `temperature_classifier` UDF#

Run the following commands to package and upload the temperature_classifier UDF deployment package to the microservice:

cd edge-ai-libraries/microservices/time-series-analytics/
rm -f temperature_classifier.tar
tar cf temperature_classifier.tar udfs/ tick_scripts/
curl -X POST http://localhost:30002/udfs/package \
  -F "file=@temperature_classifier.tar"

Activate the UDF Deployment Package#

Run the following command to apply the configuration and activate the uploaded UDF:

cd edge-ai-libraries/microservices/time-series-analytics/

curl -s -X POST http://localhost:30002/config \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d @config.json

Ingesting Temperature Data into the Time Series Analytics Microservice#

Run the following script to ingest temperature data into the Time Series Analytics Microservice:

cd edge-ai-libraries/microservices/time-series-analytics # path relative to git clone folder
python3 -m venv venv
source venv/bin/activate
pip3 install -r simulator/requirements.txt
python3 simulator/temperature_input.py --port 30002

Verify the Temperature Classifier Results#

Run following commands to see the filtered temperature results:

POD_NAME=$(kubectl get pods -n apps -o jsonpath='{.items[*].metadata.name}' | tr ' ' '\n' | grep deployment-time-series-analytics-microservice | head -n 1)
kubectl logs -f $POD_NAME -n apps

Uninstall helm charts#

helm uninstall time-series-analytics-microservice -n apps
kubectl get all -n apps # it takes few mins to have all application resources cleaned up

Troubleshooting#

Check pod details or container logs to catch any failures:

POD_NAME=$(kubectl get pods -n apps -o jsonpath='{.items[*].metadata.name}' | tr ' ' '\n' | grep deployment-time-series-analytics-microservice | head -n 1)
kubectl describe pod $POD_NAME $ -n apps # shows details of the pod
kubectl logs -f $POD_NAME -n apps | grep -i error


# Debugging UDF errors if container is not restarting and providing expected results
kubectl exec -it $POD_NAME -n apps -- /bin/bash
$ cat /tmp/log/kapacitor/kapacitor.log | grep -i error