컨테이너 모델

배포 파이프라인

소스 코드 (agent_config/, main.py, Dockerfile)
    ↓
Docker 빌드 (linux/arm64)
    ↓
ECR 푸시
    ↓
AgentCore CLI (agentcore_agent_runtime.py create)
    ↓
CodeBuild 프로젝트 → 컨테이너 빌드
    ↓
AgentCore Runtime (microVM)

Dockerfile

FROM ghcr.io/astral-sh/uv:python3.11-bookworm-slim
WORKDIR /app
ENV UV_SYSTEM_PYTHON=1 UV_COMPILE_BYTECODE=1
COPY requirements.txt requirements.txt
RUN uv pip install -r requirements.txt
RUN uv pip install aws-opentelemetry-distro>=0.10.1
ENV AWS_REGION=us-east-1 DOCKER_CONTAINER=1
RUN useradd -m -u 1000 bedrock_agentcore
USER bedrock_agentcore
EXPOSE 8080 8000
COPY . .
CMD ["opentelemetry-instrument", "python", "-m", "main"]

.bedrock_agentcore.yaml 주요 설정

agents:
  troubleshooting_agent_runtime:
    entrypoint: main.py
    platform: linux/arm64          # Graviton 기반
    container_runtime: docker
    aws:
      network_configuration:
        network_mode: PUBLIC       # 공개 네트워크
      protocol_configuration:
        server_protocol: HTTP      # HTTP 프로토콜
      observability:
        enabled: true              # OTel 추적
    authorizer_configuration:
      customJWTAuthorizer:
        discoveryUrl: "https://cognito-idp.../openid-configuration"

엔트리포인트 (main.py)

app = BedrockAgentCoreApp()

@app.entrypoint
async def invoke(payload, context):
    user_message = payload["prompt"]
    actor_id = payload["actor_id"]
    session_id = context.session_id
    # 에이전트 태스크 생성 → 스트리밍 응답

if __name__ == "__main__":
    app.run()  # HTTP 서버 시작 (포트 8080)

모듈별 배포 비교

모듈	Runtime 이름	네트워크	메모리
M1	troubleshooting_agent_runtime	PUBLIC	없음
M2	troubleshooting_agent_runtime_with_memory	PUBLIC	STM+LTM
M3 Conn	a2a_troubleshooting_agent_runtime	PUBLIC	STM_ONLY
M3 Perf	a2a_performance_agent_runtime	PUBLIC	STM_ONLY
M3 Collab	a2a_collaborator_agent_runtime	PUBLIC	있음