buildkite: opentelemetry+elastic agent overhead benchmark on a weekly basis

.buildkite/README.md

@@ -19,3 +19,12 @@ This is the Buildkite pipeline for the APM Agent java in charge of the snaposhot
 
 To view the pipeline and its configuration, click [here](https://buildkite.com/elastic/apm-agent-java-snapshot) or
 go to the definition in the `elastic/ci` repository.
+
+## opentelemetry-benchmark pipeline
+
+This is the Buildkite pipeline for the Opentelemetry Benchmark.
+
+### Pipeline Configuration
+
+To view the pipeline and its configuration, click [here](https://buildkite.com/elastic/apm-agent-java-opentelemetry-benchmark) or
+go to the definition in `opentelemetry-benchmark.yml`.

.buildkite/hooks/pre-command

@@ -8,67 +8,24 @@
 
 set -eo pipefail
 
-echo "--- Prepare vault context :vault:"
-VAULT_ROLE_ID_SECRET=$(vault read -field=role-id secret/ci/elastic-apm-agent-java/internal-ci-approle)
-export VAULT_ROLE_ID_SECRET
-
-VAULT_SECRET_ID_SECRET=$(vault read -field=secret-id secret/ci/elastic-apm-agent-java/internal-ci-approle)
-export VAULT_SECRET_ID_SECRET
-
-VAULT_ADDR=$(vault read -field=vault-url secret/ci/elastic-apm-agent-java/internal-ci-approle)
-export VAULT_ADDR
-
-# Delete the vault specific accessing the ci vault
-PREVIOUS_VAULT_TOKEN=$VAULT_TOKEN
-export PREVIOUS_VAULT_TOKEN
-unset VAULT_TOKEN
-
-echo "--- Prepare a secure temp :closed_lock_with_key:"
-# Prepare a secure temp folder not shared between other jobs to store the key ring
-export TMP_WORKSPACE=/tmp/secured
-export KEY_FILE=$TMP_WORKSPACE"/private.key"
-
-# Secure home for our keyring
-export GNUPGHOME=$TMP_WORKSPACE"/keyring"
-mkdir -p $GNUPGHOME
-chmod -R 700 $TMP_WORKSPACE
-
-echo "--- Prepare keys context :key:"
-VAULT_TOKEN=$(vault write -field=token auth/approle/login role_id="$VAULT_ROLE_ID_SECRET" secret_id="$VAULT_SECRET_ID_SECRET")
-export VAULT_TOKEN
-
-# Nexus credentials
-SERVER_USERNAME=$(vault read -field username secret/release/nexus)
-export SERVER_USERNAME
-SERVER_PASSWORD=$(vault read -field password secret/release/nexus)
-export SERVER_PASSWORD
-
-# Signing keys
-vault read -field=key secret/release/signing >$KEY_FILE
-KEYPASS_SECRET=$(vault read -field=passphrase secret/release/signing)
-export KEYPASS_SECRET
-export KEY_ID_SECRET=D88E42B4
-
-# Import the key into the keyring
-echo "$KEYPASS_SECRET" | gpg --batch --import "$KEY_FILE"
-
-echo "--- Configure git context :git:"
-# Configure the committer since the maven release requires to push changes to GitHub
-# This will help with the SLSA requirements.
-git config --global user.email "[email protected]"
-git config --global user.name "apmmachine"
-
-echo "--- Install JDK17 :java:"
-# JDK version is defined in two different locations, here and .github/workflows/maven-goal/action.yml
-JAVA_URL=https://jvm-catalog.elastic.co/jdk
-JAVA_HOME=$(pwd)/.openjdk17
-JAVA_PKG="$JAVA_URL/latest_openjdk_17_linux.tar.gz"
-curl -L --output /tmp/jdk.tar.gz "$JAVA_PKG"
-mkdir -p "$JAVA_HOME"
-tar --extract --file /tmp/jdk.tar.gz --directory "$JAVA_HOME" --strip-components 1
-
-export JAVA_HOME
-PATH=$JAVA_HOME/bin:$PATH
-export PATH
-
-java -version || true
+# Upload should not do much with the pre-command.
+if [[ "$BUILDKITE_COMMAND" =~ .*"upload".* ]]; then
+  echo "Skipped pre-command when running the Upload pipeline"
+  exit 0
+fi
+
+## TODO: change name for the opentelemetry-benchmark
+if [ "$BUILDKITE_PIPELINE_SLUG" == "apm-agent-java-load-testing" ]; then
+    source .buildkite/hooks/prepare-benchmark.sh
+fi
+
+if [ "$BUILDKITE_PIPELINE_SLUG" == "apm-agent-java-snapshot" ]; then
+    source .buildkite/hooks/prepare-release.sh
+fi
+
+if [ "$BUILDKITE_PIPELINE_SLUG" == "apm-agent-java-release" ]; then
+    source .buildkite/hooks/prepare-release.sh
+fi
+
+# Run always
+source .buildkite/hooks/prepare-common.sh

.buildkite/hooks/prepare-benchmark.sh

@@ -0,0 +1,25 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+echo "--- Prepare elasticsearch secrets :vault:"
+ES_URL_SECRET=$(vault read -field=es_url secret/ci/elastic-apm-agent-java/opentelemetry-benchmark)
+ES_USER_SECRET=$(vault read -field=es_user secret/ci/elastic-apm-agent-java/opentelemetry-benchmark)
+ES_PASS_SECRET=$(vault read -field=es_pass secret/ci/elastic-apm-agent-java/opentelemetry-benchmark)
+export ES_URL_SECRET ES_USER_SECRET ES_PASS_SECRET
+
+echo "--- Prepare github secrets :vault:"
+GITHUB_SECRET=$(vault kv get -field token "kv/ci-shared/observability-ci/github-apmmachine")
+GH_TOKEN=$GITHUB_SECRET
+export GITHUB_SECRET GH_TOKEN
+GITHUB_USERNAME=apmmachine
+export GITHUB_USERNAME
+
+echo "--- Install gh :github:"
+GH_URL=https://github.com/cli/cli/releases/download/v2.37.0/gh_2.37.0_linux_amd64.tar.gz
+GH_HOME=$(pwd)/.gh
+curl -L --output /tmp/gh.tar.gz "$GH_URL"
+mkdir -p "$GH_HOME"
+tar --extract --file /tmp/gh.tar.gz --directory "$GH_HOME" --strip-components 1
+
+PATH=$GH_HOME/bin:$PATH
+export PATH

.buildkite/hooks/prepare-common.sh

@@ -0,0 +1,17 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+echo "--- Install JDK17 :java:"
+# JDK version is defined in two different locations, here and .github/workflows/maven-goal/action.yml
+JAVA_URL=https://jvm-catalog.elastic.co/jdk
+JAVA_HOME=$(pwd)/.openjdk17
+JAVA_PKG="$JAVA_URL/latest_openjdk_17_linux.tar.gz"
+curl -L --output /tmp/jdk.tar.gz "$JAVA_PKG"
+mkdir -p "$JAVA_HOME"
+tar --extract --file /tmp/jdk.tar.gz --directory "$JAVA_HOME" --strip-components 1
+
+export JAVA_HOME
+PATH=$JAVA_HOME/bin:$PATH
+export PATH
+
+java -version || true

.buildkite/hooks/prepare-release.sh

@@ -0,0 +1,52 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+echo "--- Prepare vault context :vault:"
+VAULT_ROLE_ID_SECRET=$(vault read -field=role-id secret/ci/elastic-apm-agent-java/internal-ci-approle)
+export VAULT_ROLE_ID_SECRET
+
+VAULT_SECRET_ID_SECRET=$(vault read -field=secret-id secret/ci/elastic-apm-agent-java/internal-ci-approle)
+export VAULT_SECRET_ID_SECRET
+
+VAULT_ADDR=$(vault read -field=vault-url secret/ci/elastic-apm-agent-java/internal-ci-approle)
+export VAULT_ADDR
+
+# Delete the vault specific accessing the ci vault
+PREVIOUS_VAULT_TOKEN=$VAULT_TOKEN
+export PREVIOUS_VAULT_TOKEN
+unset VAULT_TOKEN
+
+echo "--- Prepare a secure temp :closed_lock_with_key:"
+# Prepare a secure temp folder not shared between other jobs to store the key ring
+export TMP_WORKSPACE=/tmp/secured
+export KEY_FILE=$TMP_WORKSPACE"/private.key"
+
+# Secure home for our keyring
+export GNUPGHOME=$TMP_WORKSPACE"/keyring"
+mkdir -p $GNUPGHOME
+chmod -R 700 $TMP_WORKSPACE
+
+echo "--- Prepare keys context :key:"
+VAULT_TOKEN=$(vault write -field=token auth/approle/login role_id="$VAULT_ROLE_ID_SECRET" secret_id="$VAULT_SECRET_ID_SECRET")
+export VAULT_TOKEN
+
+# Nexus credentials
+SERVER_USERNAME=$(vault read -field username secret/release/nexus)
+export SERVER_USERNAME
+SERVER_PASSWORD=$(vault read -field password secret/release/nexus)
+export SERVER_PASSWORD
+
+# Signing keys
+vault read -field=key secret/release/signing >$KEY_FILE
+KEYPASS_SECRET=$(vault read -field=passphrase secret/release/signing)
+export KEYPASS_SECRET
+export KEY_ID_SECRET=D88E42B4
+
+# Import the key into the keyring
+echo "$KEYPASS_SECRET" | gpg --batch --import "$KEY_FILE"
+
+echo "--- Configure git context :git:"
+# Configure the committer since the maven release requires to push changes to GitHub
+# This will help with the SLSA requirements.
+git config --global user.email "[email protected]"
+git config --global user.name "apmmachine"

.buildkite/load-testing.yml

@@ -1,4 +1,8 @@
-# @reakaleek: This is a place holder to create the pipeline in Buildkite. I will work on it, in a follow-up.
 steps:
-  - label: ":wave: Greetings"            # Label (with rich emojis https://ela.st/bk-emoji).
-    command: "echo 'My first pipeline!'" # Command to run (evaluated by Bash).
+  - label: "Run the microbenchmark"
+    commands: .buildkite/scripts/opentelemetry-benchmark.sh
+    agents:
+      queue: observability-microbenchmarks
+    artifact_paths:
+      - "**/build/reports/tests/test/classes/io.opentelemetry.OverheadTests.html"
+      - "output.json"

.buildkite/opentelemetry-benchmark.yml

@@ -0,0 +1,12 @@
+steps:
+  - label: "Run the opentelemetry-benchmark"
+    commands: .buildkite/scripts/opentelemetry-benchmark.sh
+    agents:
+      queue: observability-microbenchmarks
+    artifact_paths:
+      - "**/build/reports/tests/test/classes/io.opentelemetry.OverheadTests.html"
+      - "output.json"
+
+notify:
+  - slack: "#apm-agent-java"
+    if: 'build.state != "passed"'

.buildkite/scripts/opentelemetry-benchmark.sh

@@ -0,0 +1,84 @@
+#!/usr/bin/env bash
+set -eo pipefail
+
+CONTAINER_NAME=mock-apm-server
+JSON_FILE="$(pwd)/output.json"
+
+function cleanup {
+  echo "--- Tear down the environment"
+  MOCK_APM_SERVER=$(docker ps | grep $CONTAINER_NAME | awk '{print $1}')
+  docker stop $MOCK_APM_SERVER
+  docker rm $MOCK_APM_SERVER
+}
+
+trap cleanup EXIT
+
+echo "--- Download the latest elastic-apm-agent artifact"
+# run earlier so gh can use the current github repository.
+run_id=$(gh run list --branch main --status success --workflow main.yml -L 1 --json databaseId --jq '.[].databaseId')
+echo "downloading the latest artifact 'elastic-apm-agent' (using the workflow run '$run_id')"
+gh run download "$run_id" -n elastic-apm-agent
+ELASTIC_SNAPSHOT_JAR=$(ls -1 elastic-apm-agent-*.jar)
+ELASTIC_SNAPSHOT_JAR_FILE="$(pwd)/$ELASTIC_SNAPSHOT_JAR"
+echo "$ELASTIC_SNAPSHOT_JAR_FILE has been downloaded."
+gh run download "$run_id" -n apm-agent-benchmarks
+BENCHMARKS_JAR=$(ls -1 benchmarks*.jar)
+BENCHMARKS_JAR_FILE="$(pwd)/$BENCHMARKS_JAR"
+echo "$BENCHMARKS_JAR_FILE has been downloaded."
+
+echo "--- Start APM Server mock"
+git clone https://github.com/elastic/apm-mutating-webhook.git
+pushd apm-mutating-webhook/test/mock
+docker build -t $CONTAINER_NAME .
+docker run -dp 127.0.0.1:8027:8027 $CONTAINER_NAME
+popd
+
+echo "--- Build opentelemetry-java-instrumentation"
+git clone https://github.com/open-telemetry/opentelemetry-java-instrumentation.git --depth 1 --branch main
+pushd opentelemetry-java-instrumentation/
+./gradlew assemble
+
+echo "--- Customise the elastic opentelemetry java instrumentation"
+pushd benchmark-overhead
+cp "$ELASTIC_SNAPSHOT_JAR_FILE" .
+ELASTIC_SNAPSHOT_ENTRY="new Agent(\\\"elastic-snapshot\\\",\\\"latest available snapshot version from elastic main\\\",\\\"file://$PWD/$ELASTIC_SNAPSHOT_JAR\\\", java.util.List.of(\\\"-Delastic.apm.server_url=http://host.docker.internal:8027/\\\"))"
+ELASTIC_LATEST_VERSION=$(curl -s https://repo1.maven.org/maven2/co/elastic/apm/elastic-apm-agent/ | perl -ne 's/<.*?>//g; if(s/^([\d\.]+).*$/$1/){print}' | sort -V | tail -1)
+ELASTIC_LATEST_ENTRY="new Agent(\\\"elastic-latest\\\",\\\"latest available released version from elastic main\\\",\\\"https://repo1.maven.org/maven2/co/elastic/apm/elastic-apm-agent/$ELASTIC_LATEST_VERSION/elastic-apm-agent-$ELASTIC_LATEST_VERSION.jar\\\", java.util.List.of(\\\"-Delastic.apm.server_url=http://host.docker.internal:8027/\\\"))"
+ELASTIC_LATEST_ENTRY2="new Agent(\\\"elastic-async\\\",\\\"latest available released version from elastic main\\\",\\\"https://repo1.maven.org/maven2/co/elastic/apm/elastic-apm-agent/$ELASTIC_LATEST_VERSION/elastic-apm-agent-$ELASTIC_LATEST_VERSION.jar\\\", java.util.List.of(\\\"-Delastic.apm.delay_agent_premain_ms=15000\\\",\\\"-Delastic.apm.server_url=http://host.docker.internal:8027/\\\"))"
+NEW_LINE="              .withAgents(Agent.NONE, Agent.LATEST_RELEASE, Agent.LATEST_SNAPSHOT, $ELASTIC_LATEST_ENTRY, $ELASTIC_LATEST_ENTRY2, $ELASTIC_SNAPSHOT_ENTRY)"
+echo $NEW_LINE
+perl -i -ne "if (/withAgents/) {print \"$NEW_LINE\n\"}else{print}" src/test/java/io/opentelemetry/config/Configs.java
+
+echo "--- Run tests of benchmark-overhead"
+./gradlew test
+
+echo "--- Report in Buildkite"
+
+REPORT_FILE=$(pwd)/build/reports/tests/test/classes/io.opentelemetry.OverheadTests.html
+perl -ne '/Standard output/ && $on++; /\<\/pre\>/ && ($on=0);$on && s/\<.*\>//;$on && !/^\s*$/ && print' $REPORT_FILE | tee report.txt
+
+# Buildkite annotation
+if [ -n "$BUILDKITE" ]; then
+  REPORT=$(cat report.txt)
+  cat << EOF | buildkite-agent annotate --style "info" --context report
+  ### OverheadTests Report
+
+  \`\`\`
+  ${REPORT}
+  \`\`\`
+
+EOF
+fi
+
+echo "--- Generate ES docs"
+JSON_FILE="$(pwd)/output.json"
+java -cp $BENCHMARKS_JAR_FILE \
+  co.elastic.apm.agent.benchmark.ProcessOtelBenchmarkResults \
+  "$REPORT_FILE" "$JSON_FILE" "$ELASTIC_LATEST_VERSION" ./opentelemetry-javaagent.jar
+
+echo "--- Send Report"
+curl -X POST \
+  --user "${ES_USER_SECRET}:${ES_PASS_SECRET}" \
+  "${ES_URL_SECRET}/_bulk?pretty" \
+  -H "Content-Type: application/x-ndjson" \
+  --data-binary @"$JSON_FILE"

apm-agent-benchmarks/src/main/java/co/elastic/apm/agent/benchmark/ProcessOtelBenchmarkResults.java

@@ -122,7 +122,7 @@ private void writeBulkFile(String resultFilePath) throws IOException {
         final File file = new File(resultFilePath);
         final FileWriter fileWriter = new FileWriter(file);
         for (JsonNode benchmark : bechmarkResultJson) {
-            fileWriter.append("{ \"index\" : { \"_index\" : \"microbenchmarks\" } }\n");
+            fileWriter.append("{ \"index\" : { \"_index\" : \"otel-microbenchmarks\" } }\n");
             fileWriter.append(objectMapper.writer().writeValueAsString(benchmark));
             fileWriter.append("\n");
         }

catalog-info.yaml

@@ -91,3 +91,40 @@ spec:
           access_level: MANAGE_BUILD_AND_READ
         everyone:
           access_level: READ_ONLY
+
+---
+# yaml-language-server: $schema=https://gist.githubusercontent.com/elasticmachine/988b80dae436cafea07d9a4a460a011d/raw/rre.schema.json
+apiVersion: backstage.io/v1alpha1
+kind: Resource
+metadata:
+  name: buildkite-pipeline-apm-agent-java-opentelemetry-benchmark
+  description: Buildkite Opentelemetry Benchmark for apm-agent-java
+  links:
+    - title: Pipeline
+      url: https://buildkite.com/elastic/apm-agent-java-opentelemetry-benchmark
+spec:
+  type: buildkite-pipeline
+  owner: group:apm-agent-java
+  system: buildkite
+  implementation:
+    apiVersion: buildkite.elastic.dev/v1
+    kind: Pipeline
+    metadata:
+      name: apm-agent-java-opentelemetry-benchmark
+    spec:
+      repository: elastic/apm-agent-java
+      pipeline_file: ".buildkite/opentelemetry-benchmark.yml"
+      default_branch: main
+      provider_settings:
+        publish_commit_status: false
+      teams:
+        apm-agent-java:
+          access_level: MANAGE_BUILD_AND_READ
+        observablt-robots:
+          access_level: MANAGE_BUILD_AND_READ
+        everyone:
+          access_level: READ_ONLY
+      schedules:
+        Weekly Benchmark on main Branch:
+          cronline: "@weekly"
+          message: "Run the quick benchmark weekly."