mirror of
https://github.com/snapotter-hq/SnapOtter.git
synced 2026-08-03 07:46:42 +02:00
feat: unified Docker image with GPU auto-detection
- Remove VARIANT/GPU build args, single image for all platforms - amd64: nvidia/cuda base with GPU Python packages - arm64: node base with CPU Python packages - Add tini as PID 1 for proper signal handling - Replace npx tsx with node --import tsx - Split pip install into base + tool layers for better caching - Add NVIDIA_VISIBLE_DEVICES env vars for container toolkit - Suppress Python ML library log noise - Increase healthcheck start-period to 60s - Remove STIRLING_VARIANT env var - Remove lama-cleaner from pip installs
This commit is contained in:
Siddharth Kumar Sah
+67
-78
@@ -1,12 +1,9 @@
|
||||
# syntax=docker/dockerfile:1
|
||||
# ============================================
|
||||
# Stirling Image - Production Dockerfile
|
||||
# Multi-stage build for single-container deployment
|
||||
# Stirling Image - Unified Production Dockerfile
|
||||
# Single image: GPU auto-detected on amd64, CPU on arm64
|
||||
# ============================================
|
||||
|
||||
ARG VARIANT=full
|
||||
ARG GPU=false
|
||||
|
||||
# ============================================
|
||||
# Stage 1: Build the frontend (Vite + React)
|
||||
# ============================================
|
||||
@@ -39,25 +36,22 @@ RUN --mount=type=cache,id=turbo-cache,target=/app/.turbo \
|
||||
pnpm --filter @stirling-image/web build
|
||||
|
||||
# ============================================
|
||||
# Stage 2: Base image selection
|
||||
# Stage 2: Platform-specific base images
|
||||
# ============================================
|
||||
FROM node:22-bookworm AS base-cpu
|
||||
FROM nvidia/cuda:12.6.3-runtime-ubuntu24.04 AS base-gpu
|
||||
|
||||
# Select base: GPU=false -> base-cpu, GPU=true -> base-gpu
|
||||
FROM base-cpu AS production-base-false
|
||||
FROM base-gpu AS production-base-true
|
||||
FROM node:22-bookworm AS base-linux-arm64
|
||||
FROM nvidia/cuda:12.6.3-runtime-ubuntu24.04 AS base-linux-amd64
|
||||
|
||||
# ============================================
|
||||
# Stage 3: Production runtime
|
||||
# ============================================
|
||||
FROM production-base-${GPU} AS production
|
||||
ARG TARGETOS
|
||||
ARG TARGETARCH
|
||||
FROM base-${TARGETOS}-${TARGETARCH} AS production
|
||||
|
||||
ARG VARIANT
|
||||
ARG GPU
|
||||
ARG TARGETARCH
|
||||
|
||||
# Install Node.js when using CUDA base (node:22-bookworm already has it)
|
||||
RUN if [ "$GPU" = "true" ]; then \
|
||||
# Install Node.js on amd64 (CUDA base has no Node; arm64 base already has it)
|
||||
RUN if [ "$TARGETARCH" = "amd64" ]; then \
|
||||
apt-get update && apt-get install -y --no-install-recommends \
|
||||
curl ca-certificates gnupg && \
|
||||
mkdir -p /etc/apt/keyrings && \
|
||||
@@ -71,67 +65,55 @@ RUN if [ "$GPU" = "true" ]; then \
|
||||
|
||||
RUN corepack enable && corepack prepare pnpm@9.15.4 --activate
|
||||
|
||||
# System dependencies shared by all variants
|
||||
# System dependencies (all platforms)
|
||||
RUN apt-get update && apt-get install -y --no-install-recommends \
|
||||
tini \
|
||||
imagemagick \
|
||||
libraw-dev \
|
||||
potrace \
|
||||
curl \
|
||||
gosu \
|
||||
libheif-examples \
|
||||
python3 python3-pip python3-venv python3-dev \
|
||||
tesseract-ocr tesseract-ocr-eng tesseract-ocr-deu tesseract-ocr-fra tesseract-ocr-spa \
|
||||
build-essential \
|
||||
libgl1 libglib2.0-0 \
|
||||
&& rm -rf /var/lib/apt/lists/*
|
||||
|
||||
# Python/ML system dependencies (full variant only)
|
||||
RUN if [ "$VARIANT" = "full" ]; then \
|
||||
apt-get update && apt-get install -y --no-install-recommends \
|
||||
python3 python3-pip python3-venv python3-dev \
|
||||
tesseract-ocr tesseract-ocr-eng tesseract-ocr-deu tesseract-ocr-fra tesseract-ocr-spa \
|
||||
build-essential \
|
||||
libgl1 libglib2.0-0 \
|
||||
&& rm -rf /var/lib/apt/lists/* \
|
||||
# Python venv - Layer 1: Base packages (rarely change, ~3 GB)
|
||||
RUN python3 -m venv /opt/venv && \
|
||||
/opt/venv/bin/pip install --upgrade pip && \
|
||||
/opt/venv/bin/pip install \
|
||||
Pillow==11.1.0 \
|
||||
numpy==1.26.4 \
|
||||
opencv-python-headless==4.10.0.84
|
||||
|
||||
# Platform-conditional ONNX runtime
|
||||
RUN if [ "$TARGETARCH" = "amd64" ]; then \
|
||||
/opt/venv/bin/pip install onnxruntime-gpu==1.20.1 \
|
||||
; else \
|
||||
/opt/venv/bin/pip install onnxruntime==1.20.1 \
|
||||
; fi
|
||||
|
||||
# Python venv + ML packages + model weights (full variant only)
|
||||
COPY packages/ai/python/requirements.txt /tmp/requirements.txt
|
||||
COPY packages/ai/python/requirements-gpu.txt /tmp/requirements-gpu.txt
|
||||
RUN if [ "$VARIANT" = "full" ]; then \
|
||||
python3 -m venv /opt/venv && \
|
||||
/opt/venv/bin/pip install --upgrade pip && \
|
||||
if [ "$GPU" = "true" ]; then \
|
||||
/opt/venv/bin/pip install \
|
||||
Pillow numpy opencv-python-headless onnxruntime-gpu && \
|
||||
(/opt/venv/bin/pip install rembg || echo "WARNING: rembg not installed") && \
|
||||
(/opt/venv/bin/pip install realesrgan \
|
||||
--extra-index-url https://download.pytorch.org/whl/cu126 \
|
||||
|| echo "WARNING: realesrgan not installed") && \
|
||||
(/opt/venv/bin/pip install paddlepaddle-gpu paddleocr || echo "WARNING: PaddleOCR not installed") && \
|
||||
(/opt/venv/bin/pip install mediapipe || echo "WARNING: mediapipe not installed") && \
|
||||
(/opt/venv/bin/pip install lama-cleaner || echo "WARNING: lama-cleaner not installed") && \
|
||||
(/opt/venv/bin/pip install seam-carving || echo "WARNING: seam-carving not installed") \
|
||||
; else \
|
||||
/opt/venv/bin/pip install \
|
||||
Pillow numpy opencv-python-headless onnxruntime && \
|
||||
(/opt/venv/bin/pip install "rembg[cpu]" || echo "WARNING: rembg not installed") && \
|
||||
(/opt/venv/bin/pip install realesrgan || echo "WARNING: realesrgan not installed") && \
|
||||
(/opt/venv/bin/pip install paddlepaddle paddleocr || echo "WARNING: PaddleOCR not installed") && \
|
||||
(/opt/venv/bin/pip install mediapipe || echo "WARNING: mediapipe not installed") && \
|
||||
(/opt/venv/bin/pip install lama-cleaner || echo "WARNING: lama-cleaner not installed") && \
|
||||
(/opt/venv/bin/pip install seam-carving || echo "WARNING: seam-carving not installed") \
|
||||
; fi \
|
||||
; fi && rm -f /tmp/requirements.txt /tmp/requirements-gpu.txt
|
||||
# Python venv - Layer 2: Tool packages (change occasionally, ~2 GB)
|
||||
RUN if [ "$TARGETARCH" = "amd64" ]; then \
|
||||
/opt/venv/bin/pip install rembg==2.0.62 && \
|
||||
/opt/venv/bin/pip install realesrgan==0.3.0 \
|
||||
--extra-index-url https://download.pytorch.org/whl/cu126 && \
|
||||
/opt/venv/bin/pip install paddlepaddle-gpu==3.0.0 paddleocr==2.9.1 \
|
||||
; else \
|
||||
/opt/venv/bin/pip install "rembg[cpu]==2.0.62" && \
|
||||
/opt/venv/bin/pip install realesrgan==0.3.0 && \
|
||||
/opt/venv/bin/pip install paddlepaddle==3.0.0 paddleocr==2.9.1 \
|
||||
; fi
|
||||
|
||||
RUN /opt/venv/bin/pip install \
|
||||
mediapipe==0.10.21 \
|
||||
seam-carving==1.1.0
|
||||
|
||||
# Pre-download and verify all ML models
|
||||
COPY docker/download_models.py /tmp/download_models.py
|
||||
RUN if [ "$VARIANT" = "full" ]; then \
|
||||
/opt/venv/bin/python3 /tmp/download_models.py && \
|
||||
/opt/venv/bin/python3 -c "\
|
||||
try: \
|
||||
from paddleocr import PaddleOCR; \
|
||||
print('Downloading PaddleOCR models...'); \
|
||||
ocr = PaddleOCR(use_angle_cls=True, lang='en', show_log=False); \
|
||||
print('PaddleOCR models ready'); \
|
||||
except: print('PaddleOCR model pre-download skipped') \
|
||||
" 2>/dev/null || echo "WARNING: Could not pre-download PaddleOCR models" \
|
||||
; fi && rm -f /tmp/download_models.py
|
||||
RUN /opt/venv/bin/python3 /tmp/download_models.py && rm -f /tmp/download_models.py
|
||||
|
||||
WORKDIR /app
|
||||
|
||||
@@ -151,10 +133,8 @@ RUN --mount=type=cache,id=pnpm-store,target=/root/.local/share/pnpm/store/v3 \
|
||||
pnpm install --frozen-lockfile --prod
|
||||
|
||||
# Remove build tools no longer needed in production
|
||||
RUN if [ "$VARIANT" = "full" ]; then \
|
||||
apt-get purge -y --auto-remove build-essential python3-dev && \
|
||||
rm -rf /var/lib/apt/lists/* \
|
||||
; fi
|
||||
RUN apt-get purge -y --auto-remove build-essential python3-dev && \
|
||||
rm -rf /var/lib/apt/lists/*
|
||||
|
||||
# Copy source code for API (tsx runs TS directly - no build step needed)
|
||||
COPY apps/api/src ./apps/api/src
|
||||
@@ -170,9 +150,9 @@ COPY packages/ai/python ./packages/ai/python
|
||||
COPY --from=builder /app/apps/web/dist ./apps/web/dist
|
||||
|
||||
# Create required directories
|
||||
RUN mkdir -p /data /tmp/workspace
|
||||
RUN mkdir -p /data /data/files /tmp/workspace
|
||||
|
||||
# Environment defaults (matching PRD Section 16.1)
|
||||
# Environment defaults
|
||||
ENV PORT=1349 \
|
||||
NODE_ENV=production \
|
||||
AUTH_ENABLED=true \
|
||||
@@ -192,13 +172,20 @@ ENV PORT=1349 \
|
||||
MAX_BATCH_SIZE=200 \
|
||||
CONCURRENT_JOBS=3 \
|
||||
MAX_MEGAPIXELS=100 \
|
||||
RATE_LIMIT_PER_MIN=100 \
|
||||
STIRLING_VARIANT=${VARIANT}
|
||||
RATE_LIMIT_PER_MIN=100
|
||||
|
||||
# NVIDIA Container Toolkit env vars (harmless on non-GPU systems)
|
||||
ENV NVIDIA_VISIBLE_DEVICES=all \
|
||||
NVIDIA_DRIVER_CAPABILITIES=compute,utility
|
||||
|
||||
# Suppress noisy ML library output in docker logs
|
||||
ENV PYTHONWARNINGS=ignore \
|
||||
TF_CPP_MIN_LOG_LEVEL=3 \
|
||||
PADDLE_PDX_DISABLE_MODEL_SOURCE_CHECK=True
|
||||
|
||||
# Create non-root user for runtime
|
||||
RUN groupadd -r stirling && useradd -r -g stirling -d /app -s /sbin/nologin stirling
|
||||
RUN chown -R stirling:stirling /app /data /tmp/workspace && \
|
||||
([ -d /opt/venv ] && chown -R stirling:stirling /opt/venv || true)
|
||||
RUN chown -R stirling:stirling /app /data /tmp/workspace /opt/venv
|
||||
|
||||
# Entrypoint fixes volume permissions then drops to stirling via gosu
|
||||
COPY docker/entrypoint.sh /usr/local/bin/entrypoint.sh
|
||||
@@ -206,8 +193,10 @@ RUN chmod +x /usr/local/bin/entrypoint.sh
|
||||
|
||||
EXPOSE 1349
|
||||
|
||||
HEALTHCHECK --interval=30s --timeout=5s --start-period=15s --retries=3 \
|
||||
HEALTHCHECK --interval=30s --timeout=5s --start-period=60s --retries=3 \
|
||||
CMD curl -f http://localhost:1349/api/v1/health || exit 1
|
||||
|
||||
ENTRYPOINT ["entrypoint.sh"]
|
||||
CMD ["npx", "tsx", "apps/api/src/index.ts"]
|
||||
# tini as PID 1 for zombie reaping + signal forwarding
|
||||
# node --import tsx replaces npx tsx to avoid npx eating signals
|
||||
ENTRYPOINT ["tini", "--", "entrypoint.sh"]
|
||||
CMD ["node", "--import", "tsx", "apps/api/src/index.ts"]
|
||||
|
||||
Reference in New Issue
Block a user