Skip to main content

Agent Update Instructions

This document provides step-by-step instructions for updating the Pulse Agent in various environments.

Docker

  1. Pull the latest version of the image:
docker pull r.bigdataboutique.com/pulse-agent:latest
  1. Stop and remove the existing container:
docker stop pulse-agent
docker rm pulse-agent
  1. Run a new container with the same settings as before:


Kubernetes

  1. Using the Always Pull Policy with no tag or the latest tag:

    • Restart the deployment:
kubectl rollout restart deployment/pulse-agent
  1. Alternatively, delete the Pulse Agent pod:
kubectl delete pod <POD_NAME>

Kubernetes via Helm

Ensure your Helm chart is configured to use the latest tag or an Always Pull Policy.

Run:

helm upgrade --install pulse-agent <chart-repo> --set image.tag=latest

ECS

Force a new deployment:

aws ecs update-service <ECS_CLUSTER_NAME> --service-name pulse-agent --force-new-deployment
  • Replace <ECS_CLUSTER_NAME> with the name of the ECS cluster running the task.

Nomad

  1. Update the job file:
    Add or update a meta tag, such as force_update:
task "agent" {
meta {
force_update = "1"
}
...
}
  1. Rerun the job:
nomad job run pulse_agent.nomad

Cloud Run

Deploy the latest image:

gcloud run deploy pulse-agent --image r.bigdataboutique.com/pulse-agent:latest