Adding in-house version of multi-agent-travel-planner (#17)

* Adding in-house version of multi-agent-travel-planner
* run one random request once in 15m, with max timeout 10m, no retry
* fix chat llm invocations
* unify log record utils
* emit invocation logs for workflows and agents
* system_instructions fix
* temporarily change event name for agent/workflow

---------

Co-authored-by: shuning.chen <[email protected]>
Co-authored-by: Sergey Sergeev <[email protected]>

Author: 3 people

instrumentation-genai/opentelemetry-instrumentation-langchain/examples/multi_agent_travel_planner/in_house_version/Dockerfile

@@ -0,0 +1,51 @@
+# Use Python 3.12 slim image
+FROM python:3.12-slim
+
+# Set working directory
+WORKDIR /app
+
+# Install system dependencies
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    curl \
+    git \
+    && rm -rf /var/lib/apt/lists/*
+
+# Copy the local OpenTelemetry packages first
+# Note: Build context should be the repository root for this to work
+COPY util/opentelemetry-util-genai /app/util/opentelemetry-util-genai
+COPY util/opentelemetry-util-genai-evals /app/util/opentelemetry-util-genai-evals
+COPY util/opentelemetry-util-genai-evals-deepeval /app/util/opentelemetry-util-genai-evals-deepeval
+COPY util/opentelemetry-util-genai-emitters-splunk /app/util/opentelemetry-util-genai-emitters-splunk
+
+# Install the local OpenTelemetry packages
+RUN pip install --no-cache-dir -e /app/util/opentelemetry-util-genai \
+    && pip install --no-cache-dir -e /app/util/opentelemetry-util-genai-evals \
+    && pip install --no-cache-dir -e /app/util/opentelemetry-util-genai-evals-deepeval \
+    && pip install --no-cache-dir -e /app/util/opentelemetry-util-genai-emitters-splunk
+
+# Copy requirements file (for other dependencies)
+COPY instrumentation-genai/opentelemetry-instrumentation-langchain/examples/multi_agent_travel_planner/requirements.txt .
+
+# Install Python dependencies
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Copy the application code
+COPY instrumentation-genai/opentelemetry-instrumentation-langchain/examples/multi_agent_travel_planner/main.py .
+# Note: .env.example is optional, only copy if needed
+# COPY instrumentation-genai/opentelemetry-instrumentation-langchain/examples/multi_agent_travel_planner/.env.example .env.example
+
+# Create a non-root user for security
+RUN useradd --create-home --shell /bin/bash app \
+    && chown -R app:app /app
+USER app
+
+# Expose port (adjust if your app uses a different port)
+EXPOSE 8080
+
+# Health check
+HEALTHCHECK --interval=30s --timeout=10s --start-period=5s --retries=3 \
+    CMD curl -f http://localhost:8080/health || exit 1
+
+# Run the application
+CMD ["python", "main.py"]

instrumentation-genai/opentelemetry-instrumentation-langchain/examples/multi_agent_travel_planner/in_house_version/cronjob.yaml

@@ -0,0 +1,50 @@
+apiVersion: batch/v1
+kind: CronJob
+metadata:
+  name: travel-planner-load-generator
+  namespace: travel-planner
+spec:
+  schedule: "*/15 * * * *" # Every 15 minutes
+  jobTemplate:
+    spec:
+      template:
+        spec:
+          containers:
+          - name: load-generator
+            image: curlimages/curl:latest
+            command:
+            - /bin/sh
+            - -c
+            - |
+              echo "Generating travel planning requests..."
+              
+              # Simple requests without bash-specific syntax
+              destinations="Tokyo Paris NewYork Sydney Barcelona Dubai London Singapore"
+              budgets="2000 3000 1500 4000 2500"
+              durations="3 5 7 10"
+              
+              # Generate a single random request per run
+              RAND=$(date +%s)
+              dest=$(echo $destinations | cut -d' ' -f$((($RAND % 8) + 1)))
+              budget=$(echo $budgets | cut -d' ' -f$((($RAND % 5) + 1)))
+              duration=$(echo $durations | cut -d' ' -f$((($RAND % 4) + 1)))
+
+              echo "Planning trip to $dest, budget: \$$budget, duration: $duration days"
+
+              # Make single request to the travel planner service
+              curl -X POST http://travel-planner-service.travel-planner.svc.cluster.local/plan \
+                -H "Content-Type: application/json" \
+                -d "{
+                  \"destination\": \"$dest\",
+                  \"budget\": $budget,
+                  \"duration\": $duration,
+                  \"travelers\": 2,
+                  \"interests\": [\"sightseeing\", \"food\", \"culture\"]
+                }" \
+                --max-time 600 || echo "Request failed"
+              
+              echo "Load generation completed"
+          restartPolicy: OnFailure
+      backoffLimit: 3
+  successfulJobsHistoryLimit: 3
+  failedJobsHistoryLimit: 3

instrumentation-genai/opentelemetry-instrumentation-langchain/examples/multi_agent_travel_planner/in_house_version/deployment.yaml

@@ -0,0 +1,118 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: travel-planner
+  namespace: travel-planner
+  labels:
+    app: travel-planner
+spec:
+  replicas: 1
+  selector:
+    matchLabels:
+      app: travel-planner
+  template:
+    metadata:
+      labels:
+        app: travel-planner
+    spec:
+      containers:
+        - name: travel-planner
+          image: shuniche855/travel-planner:0.0.2
+          imagePullPolicy: Always
+          ports:
+            - containerPort: 8080
+              name: http
+          env:
+            # Load OpenAI API key from secret
+            - name: OPENAI_API_KEY
+              valueFrom:
+                secretKeyRef:
+                  name: openai-api-keys
+                  key: openai-api-key
+            # Service Name
+            - name: OTEL_SERVICE_NAME
+              value: "opentelemetry-python-langchain-multi-agent-in-house"
+            # Additional OTEL configuration
+            - name: OTEL_RESOURCE_ATTRIBUTES
+              value: "deployment.environment=o11y-inframon-ai"
+            - name: SPLUNK_OTEL_AGENT
+              valueFrom:
+                fieldRef:
+                  fieldPath: status.hostIP
+            - name: OTEL_EXPORTER_OTLP_ENDPOINT
+              value: "http://$(SPLUNK_OTEL_AGENT):4317"
+            - name: OTEL_EXPORTER_OTLP_PROTOCOL
+              value: "grpc"
+            - name: HOME
+              value: "/tmp"
+            - name: OTEL_EXPORTER_OTLP_METRICS_TEMPORALITY_PREFERENCE
+              value: "DELTA"
+            - name: OTEL_PYTHON_LOGGING_AUTO_INSTRUMENTATION_ENABLED
+              value: "true"
+            - name: OTEL_PYTHON_EXCLUDED_URLS
+              value: "^(https?://)?[^/]+(/health)?$"
+            - name: OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT
+              value: "true"
+            - name: OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT_MODE
+              value: "SPAN_AND_EVENT"
+            - name: OTEL_INSTRUMENTATION_GENAI_EVALS_RESULTS_AGGREGATION
+              value: "true"
+            - name: OTEL_INSTRUMENTATION_GENAI_EMITTERS
+              value: "span_metric_event,splunk"
+            - name: OTEL_INSTRUMENTATION_GENAI_EMITTERS_EVALUATION
+              value: "replace-category:SplunkEvaluationResults"
+            - name: OTEL_GENAI_EVAL_DEBUG_SKIPS
+              value: "true"
+            - name: OTEL_GENAI_EVAL_DEBUG_EACH
+              value: "true"
+            - name: OTEL_INSTRUMENTATION_GENAI_DEBUG
+              value: "true"
+            - name: SPLUNK_PROFILER_ENABLED
+              value: "true"
+            # Set evaluation wait time to 10 seconds (short enough to avoid health check timeout)
+            - name: EVAL_WAIT_SECONDS
+              value: "10"
+          resources:
+            requests:
+              memory: "512Mi"
+              cpu: "250m"
+            limits:
+              memory: "2Gi"
+              cpu: "1000m"
+          securityContext:
+            allowPrivilegeEscalation: false
+            runAsNonRoot: true
+            runAsUser: 1000
+            capabilities:
+              drop:
+                - ALL
+            readOnlyRootFilesystem: false
+          livenessProbe:
+            httpGet:
+              path: /health
+              port: 8080
+            initialDelaySeconds: 30
+            periodSeconds: 10
+            timeoutSeconds: 5
+            failureThreshold: 6
+          readinessProbe:
+            httpGet:
+              path: /health
+              port: 8080
+            initialDelaySeconds: 5
+            periodSeconds: 5
+      restartPolicy: Always
+---
+apiVersion: v1
+kind: Service
+metadata:
+  name: travel-planner-service
+  namespace: travel-planner
+spec:
+  selector:
+    app: travel-planner
+  ports:
+    - name: http
+      port: 80
+      targetPort: 8080
+  type: ClusterIP