Prometheus E2E benchmarking tool
Switch branches/tags
Nothing to show
Clone or download
sipian and krasi-georgiev Update grafana graphs (#158)
Separate timing , WAL fsync/latency graphs
Divide latency by _count to get average per query / request
Signed-off-by: Harsh Agarwal <cs15btech11019@iith.ac.in>
Latest commit 7275029 Sep 20, 2018

README.md

Automated Prometheus E2E testing and benchmarking.

Prombench Design

It runs with Prow CI on a Google Kubernetes Engine Cluster. It is designed to support adding more k8s providers.

How to trigger tests on Github

A Prometheus maintainer can comment as follows to benchmark a PR:

  • /benchmark equivalent to /benchmark master - benchmark PR with the master branch.
  • /benchmark 2.4.0 - benchmark the PR with the 2.4.0 release(could be any release). Don't prepend v to the release version.

To cancel benchmarking, a mantainer should comment /benchmark cancel.

Prerequisites

  • Create a new Google cloud project - prometheus-ci

  • Create a Service Account on GKE with role Kubernetes Engine Service Agent & Kubernetes Engine Admin and download the json file.

  • Generate a github auth token that will be used to authenticate when sending requests to the github api.

  • Set some env variable which will be used in the commands below.

    • Note: The #GCLOUD_SERVICEACCOUNT_CLIENTID is used to grant cluster-admin-rights to the service-account which needs to create RBAC roles. The service-account is used by the prombench tool when managing the cluster for each job.
export PROJECT_ID=prometheus-ci 
export CLUSTER_NAME=prow
export ZONE=europe-west3-a
export AUTH_FILE=<path to service-account.json>
export GITHUB_ORG=prometheus
export GITHUB_REPO=prometheus
export GRAFANA_ADMIN_PASSWORD=$(openssl rand -hex 20)
export HMAC_TOKEN=$(openssl rand -hex 20)
export OAUTH_TOKEN=***Replace with the generated token from github***
export GCLOUD_SERVICEACCOUNT_CLIENTID=<client_id from the service-account.json>
  • Add a github webhook where to send the events.
    • Content Type: json

    • Send: Issue comments,Pull requests

    • Secret: echo $HMAC_TOKEN

    • Payload URL: http://prombench.prometheus.io/hook

      • Note: The ip DNS record for prombench.prometheus.io will be added once we get it from the ingress deployment in the following steps.

Prow Setup.

  • Create the main k8s cluster to deploy the Prow components.
./prombench gke cluster create -a $AUTH_FILE -v PROJECT_ID:$PROJECT_ID \
-v ZONE:$ZONE -v CLUSTER_NAME:$CLUSTER_NAME -f components/prow/cluster.yaml
  • Add all required tokens as k8s secrets.
    • hmac is used when verifying requests from github.
    • oauth is used when sending requests to the github api.
    • gke auth is used when scaling up and down the cluster.
./prombench gke resource apply -a $AUTH_FILE -v ZONE:$ZONE -v CLUSTER_NAME:$CLUSTER_NAME \
-f components/prow/manifests/secrets.yaml \
-v HMAC_TOKEN="$(printf $HMAC_TOKEN | base64 -w 0)" \
-v OAUTH_TOKEN="$(printf $OAUTH_TOKEN | base64 -w 0)" \
-v GKE_AUTH="$(cat $AUTH_FILE | base64 -w 0)"

./prombench gke resource apply -a $AUTH_FILE -v ZONE:$ZONE -v CLUSTER_NAME:$CLUSTER_NAME \
-v GCLOUD_SERVICEACCOUNT_CLIENTID:$GCLOUD_SERVICEACCOUNT_CLIENTID \
-f components/prow/manifests/rbac.yaml -f components/prow/manifests/nginx-controller.yaml

Get the ingress ip and use it to set the DNS ip record for prombench.prometheus.io.

kubectl get ingress ing -o go-template='{{ range .status.loadBalancer.ingress}}{{.ip}}{{ end }}'
  • Deploy all internal prow components

    • Note: Long term plans are to use the prombench cli tool to deploy and manage everything, but at the moment the k8s golang client doesn't support CustomResourceDefinition objects so for those it uses kubectl.
// Generate auth config so we can use kubectl.
gcloud container clusters get-credentials $CLUSTER_NAME --zone=$ZONE
kubectl apply -f components/prow/manifests/prow_internals_1.yaml

./prombench gke resource apply -a $AUTH_FILE -v PROJECT_ID:$PROJECT_ID \
-v ZONE:$ZONE -v CLUSTER_NAME:$CLUSTER_NAME \
-v GITHUB_ORG:$GITHUB_ORG -v GITHUB_REPO:$GITHUB_REPO \
-f components/prow/manifests/prow_internals_2.yaml
  • Deploy the components that will collect and display the results.
export INGRESS_IP=$(kubectl get ingress ing -o go-template='{{ range .status.loadBalancer.ingress}}{{.ip}}{{ end }}')

./prombench gke resource apply -a $AUTH_FILE -v PROJECT_ID:$PROJECT_ID \
-v ZONE:$ZONE -v CLUSTER_NAME:$CLUSTER_NAME -v INGRESS_IP:$INGRESS_IP \
-v GRAFANA_ADMIN_PASSWORD:$GRAFANA_ADMIN_PASSWORD -f components/prombench/manifests/results

The services will be accessible at: