gke.md

Prombench in GKE

Run Prombench tests in Google Kubernetes Engine (GKE).

Table of Contents

1. Setup Prombench
   • Create the Main Node
   • Deploy Monitoring Components
2. Usage
   • Start a Benchmarking Test Manually
   • Stopping a Benchmarking Test Manually

Setup Prombench

1. Create the Main Node

---

1. Create a New Project: Start by creating a new project on Google Cloud.

2. Create a Service Account:

   • Create a Service Account on GKE with the roles:
     • Kubernetes Engine Service Agent
     • Kubernetes Engine Admin
   • If using the gcloud CLI, add the following roles to the GCP Service Account:
     • roles/container.admin
     • roles/iam.serviceAccountUser
   • Download the JSON file for the service account.

3. Set Environment Variables and Deploy the Cluster:

   export GKE_PROJECT_ID=<google-cloud project-id>
   export CLUSTER_NAME=prombench
   export ZONE=us-east1-b
   export AUTH_FILE=<path to service-account.json>
   export PROVIDER=gke
   
   make cluster_create

2. Deploy Monitoring Components

---

Note: These components are responsible for collecting, monitoring, and displaying test results and logs.

1. Optional GitHub Integration:

   • If used with GitHub integration, generate a GitHub auth token:
     • Login with the Prombot account and generate a new auth token.
     • Required permissions: public_repo, read:org, write:discussion.

   export SERVICEACCOUNT_CLIENT_EMAIL=<client-email present in service-account.json>
   export GRAFANA_ADMIN_PASSWORD=password
   export DOMAIN_NAME=prombench.prometheus.io # Can be set to any other custom domain or an empty string if not used with the GitHub integration.
   export OAUTH_TOKEN=<generated token from GitHub or set to an empty string " ">
   export WH_SECRET=<GitHub webhook secret>
   export GITHUB_ORG=prometheus
   export GITHUB_REPO=prometheus

2. Deploy the Monitoring Components:

   • This step will deploy the nginx-ingress-controller, Prometheus-Meta, Loki, Grafana, Alertmanager, and GitHub Notifier.

   make cluster_resource_apply

3. Configure DNS:

   • The output will display the ingress IP. Use this IP to point the domain name.
   • Set the A record for <DOMAIN_NAME> to point to the nginx-ingress-controller IP address.

4. Access the Services:

   • Grafana: http://<DOMAIN_NAME>/grafana
   • Prometheus: http://<DOMAIN_NAME>/prometheus-meta
   • Logs: http://<DOMAIN_NAME>/grafana/explore
   • Profiles: http://<DOMAIN_NAME>/profiles

Usage

1. Start a Benchmarking Test Manually

---

1. Set the Environment Variables:

   export RELEASE=<master/main or any Prometheus release (e.g., v2.3.0)>
   export PR_NUMBER=<PR to benchmark against the selected $RELEASE>

2. Create Node Pools for Kubernetes Objects:

   make node_create

3. Deploy the Kubernetes Objects:

   make resource_apply

2. Stopping a Benchmarking Test Manually

---

1. Set the Environment Variables:

   export GKE_PROJECT_ID=<google-cloud project-id>
   export CLUSTER_NAME=prombench
   export ZONE=us-east1-b
   export AUTH_FILE=<path to service-account.json>
   export PROVIDER=gke
   
   export PR_NUMBER=<PR to benchmark against the selected $RELEASE>

2. Delete Node Pools (Keeping the Main Node Intact):

   make clean

3. Delete Everything (Complete Teardown):

   make cluster_delete