Upgrade OpenTelemetry Demo to 3.0.0 across all deployment modes

[kylehounslow]

Summary

OpenTelemetry Demo latest version 3.0.0 changes the core service set, load generator, and several service configs. This repository vendors the demo across three deployment modes; all three need to move to 3.0.0 as per upstream issue: open-telemetry/opentelemetry-demo#3720

Scope

• 1. Docker Compose (docker-compose.otel-demo.yml, .env, vendored config under docker-compose/opentelemetry-demo/, docker-compose/opensearch-dashboards/init/, docker-compose/prometheus/rules-otel-demo/)
• 2. Kubernetes / Helm (charts/observability-stack, the opentelemetry-demo subchart, and the charts/observability-stack/files dashboard and alert-rule copies)
• 3. AWS CDK and CLI installer (aws/cdk, aws/cli-installer)

Tasks

Docker Compose

• Bump IMAGE_VERSION/DEMO_VERSION to 3.0.0 and the flagd/postgres/valkey image pins to the 3.0 set.
• Drop services removed from the 3.0 core demo: accounting, fraud-detection, kafka, llm, product-reviews. Add telemetry-docs.
• Replace the Locust load generator with k6 (LOAD_GENERATOR_VUS, K6_TARGET_URL).
• Rename postgresql to astronomy-db; update init.sql to provision astronomy_db with astronomy_user/monitoring_user.
• Mount the new product-catalog/otel-config.yml.
• Refresh demo.flagd.json to the 3.0 flag set.
• Set the frontend-proxy backend env vars (see Notes).

Kubernetes / Helm

• Bump the opentelemetry-demo subchart (currently 0.40.8) to the release that ships demo 3.0.0.
• Update the charts/observability-stack/files astronomy dashboards and otel-demo alert rules to the 3.0 service set, in lockstep with the subchart bump. These feed the init-dashboards and cortex-rules configmaps, so shipping them ahead of the subchart bump leaves the Helm path half-updated.
• Set the frontend-proxy backend env vars in the subchart values.

AWS CDK / CLI

• Propagate the 3.0.0 image version and service-set changes through aws/cdk and aws/cli-installer.
• Set the frontend-proxy backend env vars in the CDK/CLI-rendered deployment.

Notes (verified on Docker Compose)

• frontend-proxy backend vars: the 3.0.0 frontend-proxy image's envoy bootstrap defines STRICT_DNS clusters for chatbot, firepit, and opamp-server. A deployment that omits those services must still set CHATBOT_HOST/CHATBOT_PORT, FIREPIT_HOST/FIREPIT_PORT, and OPAMP_SERVER_HOST/OPAMP_SERVER_UI_PORT (upstream 3.0.0 defaults). Without them the socket addresses are empty, envoy proto validation fails, and the proxy crash-loops, which also blocks any service depending on frontend-proxy being healthy (for example the load generator). Template rendering (docker compose config) does not catch this, because the addresses are assembled inside the image at boot.
• Dashboards: remove the product-reviews panels from the astronomy-shop dashboard; drop fraud-detection and product-reviews from the service-telemetry dashboard variable list. Repoint the product-catalog throughput panel from the removed npgsql (.NET) counters to the Go client metric db_client_operation_duration_seconds_count.
• Alert rules: drop product-reviews and accounting from the otel-demo alert-rule service matchers.
• Validation: the Docker Compose 3.0.0 deployment brought up all core demo services plus the stack collector healthy, with traces, metrics, and logs flowing under k6 load.