Automating initialization of on-demand self-hosted CNCF CIL runner

.github/workflows/configurable-benchmark-test-self-hosted.yaml

@@ -0,0 +1,133 @@
+name: Configurable Benchmark Test on Self-hosted Runner
+on: 
+  workflow_dispatch:
+    inputs:
+      profile_name:
+        description: "performance profile to use"
+        required: false
+      profile_filename:
+        description: "test configuration file"
+        required: false
+      service_mesh:
+        type: choice
+        required: false
+        description: "service mesh being tested"
+        options:
+          - istio
+          - linkerd
+      load_generator:
+        type: choice
+        required: false
+        description: "load generator to run tests with"
+        options:
+          - fortio
+          - wrk2
+          - nighthawk
+
+jobs:
+  start-runner:
+    name: Start self-hosted CNCF CIL runner
+    runs-on: ubuntu-latest
+    if: ${{ github.event_name == 'workflow_dispatch' }}
+    outputs:
+      hostname: ${{ steps.start-cil-runner.outputs.hostname }}
+      label: ${{ steps.start-cil-runner.outputs.label }}
+      device_id: ${{ steps.start-cil-runner.outputs.device_id }}
+    steps:
+      - name: Checkout Code
+        uses: actions/checkout@v2
+
+      - name: Configure CNCF CIL credentials
+        run: |
+          chmod +x .github/workflows/scripts/self-hosted-credentails.sh 
+          .github/workflows/scripts/self-hosted-credentails.sh ${{ secrets.CNCF_CIL_TOKEN }}
+        shell: bash
+
+      - name: Create registration token for CNCF CIL runner
+        id: getRegToken
+        run: |
+          reg_token=$(curl -s -X POST -H "Accept: application/vnd.github.v3+json" \
+          -H 'Authorization: token ${{ secrets.PAT }}' \
+          https://api.github.com/repos/${{github.repository}}/actions/runners/registration-token | jq -r .token)
+          echo REG_TOKEN=$reg_token >> $GITHUB_ENV
+          echo REPOSITORY=${{github.repository}} >> $GITHUB_ENV
+        shell: bash
+
+      - name: Start CNCF CIL runner
+        id: start-cil-runner
+        run: |
+          chmod +x .github/workflows/scripts/start-cil-runner.sh
+          .github/workflows/scripts/start-cil-runner.sh ${{ secrets.cncf_cil_token }} ${{ github.event.inputs.service_mesh }}-${{ github.event.inputs.load_generator }}
+        shell: bash
+
+  run-benchmarks:
+    name: Run the configurable benchmarks on the runner
+    needs: 
+       - start-runner # required to start the main job when the runner is ready
+    runs-on: ${{ needs.start-runner.outputs.label }} # run the job on the newly created runner
+    steps:
+      - name: Install dependencies
+        run: |
+          echo "Current user: $(whoami)"
+          echo "Installing kubectl..."
+          curl -LO https://dl.k8s.io/release/v1.23.2/bin/linux/amd64/kubectl 
+          sudo install -o smp -g smp -m 0755 kubectl /usr/local/bin/kubectl
+          echo "Installing docker..."
+          sudo apt update -y
+          sudo apt install -y jq unzip apt-transport-https ca-certificates software-properties-common
+          curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add -
+          sudo add-apt-repository "deb [arch=amd64] https://download.docker.com/linux/ubuntu focal stable"
+          sudo apt-cache policy docker-ce
+          sudo apt install -y docker-ce
+          sudo systemctl status docker
+
+      - name: Setup Kubernetes
+        uses: manusa/[email protected]
+        with:
+          minikube version: 'v1.23.2'
+          kubernetes version: 'v1.23.2'
+          driver: docker
+
+      - name: Checkout Code
+        uses: actions/checkout@v2
+
+      - name: Install Service Mesh and Deploy Application
+        run: |
+          chmod +x .github/workflows/scripts/${{ github.event.inputs.service_mesh }}_deploy.sh
+          .github/workflows/scripts/${{ github.event.inputs.service_mesh }}_deploy.sh
+        shell: bash
+
+      - name: Run Benchmark Tests
+        uses: layer5io/meshery-smp-action@self-hosted
+        with:
+          provider_token: ${{ secrets.MESHERY_TOKEN }}
+          platform: docker
+          profile_name: ${{ github.event.inputs.profile_name }}
+          profile_filename: ${{ github.event.inputs.profile_filename }}
+          endpoint_url: ${{env.ENDPOINT_URL}}
+          service_mesh: ${{env.SERVICE_MESH}}
+          load_generator: ${{ github.event.inputs.load_generator }}
+          test_name: '${{ github.event.inputs.service_mesh }}-${{ github.event.inputs.load_generator }}-${{ github.event.inputs.profile_filename }}${{ github.event.inputs.profile_name }}'
+
+  stop-runner:
+    name: Stop self-hosted runner
+    needs:
+      - start-runner # required to get output from the start-runner job
+      - run-benchmarks # required to wait when the main job is done
+    runs-on: ubuntu-latest
+    if: ${{ always() }} # required to stop the runner even if the error happened in the previous jobs
+    steps:
+      - name: Checkout Code
+        uses: actions/checkout@v2
+
+      - name: Stop CNCF CIL runner
+        run: |
+          chmod +x .github/workflows/scripts/stop-cil-runner.sh
+          .github/workflows/scripts/stop-cil-runner.sh ${{ secrets.cncf_cil_token }} ${{ needs.start-runner.outputs.device_id }} ${{ needs.start-runner.outputs.hostname }}
+        shell: bash
+
+      - name: Remove CNCF CIL runner from github repository
+        run: |
+          runner_id=$(curl -s -H 'Authorization: token ${{ secrets.PAT }}' -H "Accept: application/vnd.github.v3+json" https://api.github.com/repos/${{github.repository}}/actions/runners | jq '.runners[] | select(.name == "${{ needs.start-runner.outputs.hostname }}") | {id}' | jq .id)
+          curl -X DELETE -H 'Authorization: token ${{ secrets.PAT }}' -H "Accept: application/vnd.github.v3+json" https://api.github.com/repos/${{github.repository}}/actions/runners/$runner_id
+        shell: bash

.github/workflows/configurable-benchmark-test.yaml

@@ -31,7 +31,7 @@ on:
 jobs:
   manual-test:
     name: Configurable Benchmark Test
-    runs-on: self-hosted
+    runs-on: ubuntu-latest
     if: ${{ github.event_name == 'workflow_dispatch' }}
     steps:
       - name: Setup Kubernetes
@@ -55,7 +55,7 @@ jobs:
         uses: layer5io/meshery-smp-action@self-hosted
         with:
           provider_token: ${{ secrets.MESHERY_TOKEN }}
-          platform: docker
+          platform: kubernetes
           profile_name: ${{ github.event.inputs.profile_name }}
           profile_filename: ${{ github.event.inputs.profile_filename }}
           endpoint_url: ${{env.ENDPOINT_URL}}

.github/workflows/scripts/self-hosted-credentails.sh

@@ -0,0 +1,14 @@
+#!/usr/bin/env bash
+
+# This script is used to configure and verify the token of self-hosted runner
+# Token must be set as a github repo secret named "CNCF_CIL_TOKEN"
+
+token=$1
+
+# https://metal.equinix.com/developers/api/authentication/#authentication
+result=$(curl -I -s -w %{http_code} -o /dev/null -H "X-Auth-Token: $token" https://api.equinix.com/metal/v1)
+if [[ $result != "200" ]]; then
+    echo "ERROR: Failed to authenticate the CNCF CIL token"
+    exit 1
+fi
+echo "Authenticate CNCF CIL token sucessfully!"

.github/workflows/scripts/start-cil-runner.sh

@@ -0,0 +1,49 @@
+#!/usr/bin/env bash
+
+# This script is used to start a CNCF CIL runner 
+
+token=$1
+hostname=$2
+
+# Generate random number from datastamp as the hostname of runner
+label=$(date +%N)
+
+hostname="$hostname-$label"
+echo "Creating CNCF CIL machine: $hostname..."
+
+# Use user_data_scripts to register the CNCF CIL runner as a self-hosted runner
+user_data_scripts="#cloud-config\nusers:\n    - default\n    - name: smp\n      groups: sudo, docker\n      sudo: ALL=(ALL) NOPASSWD:ALL\n      lock_passwd: true\nruncmd:\n    - [runuser, -l, smp, -c, \'mkdir actions-runner && cd actions-runner\']\n    - [runuser, -l, smp, -c, \'curl -o actions-runner-linux-x64-2.287.1.tar.gz -L https://github.com/actions/runner/releases/download/v2.287.1/actions-runner-linux-x64-2.287.1.tar.gz\']\n    - [runuser, -l, smp, -c, \'tar xzf ./actions-runner-linux-x64-2.287.1.tar.gz\']\n    - [runuser, -l, smp, -c, \'export RUNNER_ALLOW_RUNASROOT=1\']\n    - [runuser, -l, smp, -c, \'./config.sh --url https://github.com/$REPOSITORY --token $REG_TOKEN --labels $hostname >> github-action-registeration.log\']\n    - [runuser, -l, smp, -c, \'./run.sh >> github-action-registeration.log\']"
+
+# TODO: the options "operating_system", "facility", "plan" are hardcoded now, we should make them configurable
+# https://metal.equinix.com/developers/api/devices/#devices-createdevice
+device_id=$(curl -X POST -H "X-Auth-Token: $token" -s -H "Content-Type: application/json" \
+-d '{"operating_system": "ubuntu_20_04", "facility": "da11", "plan": "c3.small.x86", "hostname": "'"${hostname}"'", "userdata": "'"${user_data_scripts}"'"}' \
+https://api.equinix.com/metal/v1/projects/96a9d336-541b-42f7-9827-d845010da550/devices | jq -r .id)
+if [[ -z $device_id ]]; then
+    echo "ERROR: Failed to create CNCF CIL machine: $hostname..."
+    exit 1
+fi
+
+# Wait 10 minutes until the machine is running 
+echo "Waiting for $hostname to run..."
+n=0
+while [[ $n -le 10 ]]
+do
+    if [[ $n -eq 10 ]]; then
+        echo "Waiting too long for $hostname to start, exiting..."
+        exit 1 
+    fi
+    sleep 1m
+    state=$(curl -s -H "X-Auth-Token: $token" https://api.equinix.com/metal/v1/devices/$device_id | jq -r .state)
+    if [[ $state == "active" ]]; then
+        echo "$hostname successfully created!"
+        break
+    fi
+    echo "Still waiting..."
+    let n++
+done
+
+# Set the outputs
+echo "::set-output name=hostname::$hostname"
+echo "::set-output name=label::$hostname"
+echo "::set-output name=device_id::$device_id"

.github/workflows/scripts/stop-cil-runner.sh

@@ -0,0 +1,17 @@
+#!/usr/bin/env bash
+
+# This script is used to start a CNCF CIL runner
+
+token=$1
+device_id=$2
+hostname=$3
+
+echo "Removing CNCF CIL machine: $hostname..."
+
+# https://metal.equinix.com/developers/api/devices/#devices-deletedevice
+remove_cil_result=$(curl -X DELETE -I -s -w %{http_code} -o /dev/null -H "X-Auth-Token: $token" https://api.equinix.com/metal/v1/devices/$device_id)
+
+if [[ $remove_cil_result != "204" ]]; then
+    echo "ERROR: Failed to remove CNCF CIL machine: $hostname."
+    exit 1
+fi

meshery.sh

@@ -35,7 +35,7 @@ main() {
 	kubectl config view --minify --flatten > ~/minified_config
 	mv ~/minified_config ~/.kube/config
 
-  curl -L https://git.io/meshery | PLATFORM=$PLATFORM bash -
+  	curl -L https://git.io/meshery | sudo PLATFORM=$PLATFORM bash - &
 
 	sleep 60
 }