feat: AE-1018 Add support for new serverless runtime

Dockerfile-load-balancer-sls

@@ -0,0 +1,45 @@
+# Load Balancer SLS Dockerfile - Dual-capability serverless runtime
+# Supports both remote execution and HTTP endpoints
+
+# Use Python 3.11 base image with CUDA support
+FROM pytorch/pytorch:2.2.0-cuda12.1-cudnn8-runtime AS builder
+
+# Set working directory
+WORKDIR /app
+
+# Install system dependencies
+RUN apt-get update && apt-get install -y \
+    curl \
+    wget \
+    git \
+    build-essential \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install uv for fast Python package management
+COPY --from=ghcr.io/astral-sh/uv:latest /uv /uvx /bin/
+
+# Copy Python project files
+COPY pyproject.toml uv.lock ./
+COPY src/ ./src/
+
+# Install Python dependencies
+RUN uv sync --frozen
+
+# Set Python path
+ENV PYTHONPATH="/app/src:$PYTHONPATH"
+
+# Environment variables for Load Balancer SLS
+ENV RUNTIME_MODE="dual"
+ENV ENABLE_HTTP_SERVER="true"
+ENV ENABLE_REMOTE_EXECUTION="true"
+ENV PORT="8000"
+
+# Expose port for HTTP endpoints
+EXPOSE 8000
+
+# Health check
+HEALTHCHECK --interval=30s --timeout=10s --start-period=5s --retries=3 \
+    CMD curl -f http://localhost:8000/health || exit 1
+
+# Default command - can run in both serverless and HTTP modes
+CMD ["uv", "run", "python", "src/load_balancer_sls_handler.py"]

Makefile

@@ -1,7 +1,9 @@
 IMAGE = runpod/tetra-rp
+LOAD_BALANCER_SLS_IMAGE = mwiki/worker-tetra-load-balancer-sls
 TAG = local
 FULL_IMAGE = $(IMAGE):$(TAG)
 FULL_IMAGE_CPU = $(IMAGE)-cpu:$(TAG)
+FULL_IMAGE_LOAD_BALANCER_SLS = $(LOAD_BALANCER_SLS_IMAGE):$(TAG)
 
 .PHONY: setup help
 
@@ -35,9 +37,10 @@ setup: dev # Initialize project, sync deps, update submodules
 	git submodule update --remote --merge
 	cp tetra-rp/src/tetra_rp/protos/remote_execution.py src/
 
-build: # Build both GPU and CPU Docker images
+build: # Build all Docker images (GPU, CPU, Load Balancer SLS)
 	make build-gpu
 	make build-cpu
+	make build-load-balancer-sls
 
 build-gpu: setup # Build GPU Docker image (linux/amd64)
 	docker buildx build \
@@ -52,6 +55,13 @@ build-cpu: setup # Build CPU-only Docker image (linux/amd64)
 	-t $(FULL_IMAGE_CPU) \
 	. --load
 
+build-load-balancer-sls: setup # Build Load Balancer SLS dual-capability image (linux/amd64)
+	docker buildx build \
+	--platform linux/amd64 \
+	-f Dockerfile-load-balancer-sls \
+	-t $(FULL_IMAGE_LOAD_BALANCER_SLS) \
+	. --load
+
 # Test commands
 test: # Run all tests
 	uv run pytest tests/ -v
@@ -71,6 +81,9 @@ test-fast: # Run tests with fast-fail mode
 test-handler: # Test handler locally with all test_*.json files
 	cd src && ./test-handler.sh
 
+test-load-balancer-sls: build-load-balancer-sls # Test Load Balancer SLS container locally
+	docker run --rm -p 8000:8000 $(FULL_IMAGE_LOAD_BALANCER_SLS)
+
 # Smoke Tests (local on Mac OS)
 
 smoketest-macos-build: setup # Build CPU-only Mac OS Docker image (macos/arm64)

pyproject.toml

@@ -9,6 +9,9 @@ dependencies = [
     "pydantic>=2.11.4",
     "requests>=2.25.0",
     "runpod",
+    "fastapi>=0.104.0",
+    "uvicorn[standard]>=0.24.0",
+    "aiohttp>=3.9.0",
     "hf_transfer>=0.1.0",
     "huggingface_hub>=0.32.0",
 ]