File size: 12,363 Bytes
b6ae7b8 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 | #!/bin/bash
#
# Stack 2.9 Turnkey Deployment Script
# One-command deployment for Local, RunPod, and VastAI platforms
#
# Usage:
# ./deploy.sh [platform] [options]
#
# Platforms:
# local - Deploy locally with docker-compose
# runpod - Deploy to RunPod (requires runpodctl)
# vastai - Deploy to Vast.ai (requires vastai)
# kubernetes - Deploy to Kubernetes cluster
#
# Examples:
# ./deploy.sh local --model TheBloke/Llama-2-7B-Chat-AWQ
# ./deploy.sh runpod --gpu A100-40GB
# ./deploy.sh kubernetes --namespace inference
#
set -euo pipefail
# Colors for output
RED='\033[0;31m'
GREEN='\033[0;32m'
YELLOW='\033[1;33m'
BLUE='\033[0;34m'
cyan='\033[0;36m'
NC='\033[0m' # No Color
# Default values
PLATFORM="${1:-local}"
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
COMPOSE_FILE="${SCRIPT_DIR}/docker-compose.yaml"
IMAGE_NAME="stack-2.9-server"
VERSION="2.9.0"
BUILD_ARGS=""
MODEL_ID=""
HF_TOKEN=""
# Print banner
echo -e "${BLUE}"
echo "ββββββββββββββββββββββββββββββββββββββββββββββββββββββ"
echo "β Stack 2.9 Deployment Tool v${VERSION} β"
echo "β Turnkey LLM Inference Server Deployment β"
echo "ββββββββββββββββββββββββββββββββββββββββββββββββββββββ"
echo -e "${NC}"
# Parse arguments
parse_args() {
shift
while [[ $# -gt 0 ]]; do
case $1 in
--model)
MODEL_ID="$2"
shift 2
;;
--token)
HF_TOKEN="$2"
shift 2
;;
--gpu)
GPU_TYPE="$2"
shift 2
;;
--namespace)
K8S_NAMESPACE="$2"
shift 2
;;
--help)
show_help
exit 0
;;
*)
echo -e "${RED}Unknown option: $1${NC}"
show_help
exit 1
;;
esac
done
}
show_help() {
echo "Usage: $0 [platform] [options]"
echo ""
echo "Platforms:"
echo " local Deploy with docker-compose (default)"
echo " runpod Deploy to RunPod.io"
echo " vastai Deploy to Vast.ai"
echo " kubernetes Deploy to Kubernetes"
echo ""
echo "Options:"
echo " --model ID Hugging Face model ID (default: TheBloke/Llama-2-7B-Chat-AWQ)"
echo " --token TOKEN Hugging Face token for gated models"
echo " --gpu TYPE GPU type for cloud deployments"
echo " --namespace NS Kubernetes namespace (default: default)"
echo " --help Show this help message"
echo ""
echo "Examples:"
echo " $0 local --model mistralai/Mistral-7B-Instruct-v0.2"
echo " $0 runpod --gpu A100-40GB"
echo " $0 kubernetes --namespace inference"
}
# Platform detection
detect_platform() {
if command -v nvidia-smi &> /dev/null; then
echo -e "${GREEN}β NVIDIA GPU detected${NC}"
HAS_GPU=true
else
echo -e "${YELLOW}β No NVIDIA GPU detected - CPU-only mode${NC}"
HAS_GPU=false
fi
if command -v docker &> /dev/null && command -v docker-compose &> /dev/null; then
echo -e "${GREEN}β Docker & Docker Compose available${NC}"
HAS_DOCKER=true
else
echo -e "${RED}β Docker not available${NC}"
HAS_DOCKER=false
fi
}
# Build Docker image
build_image() {
echo -e "\n${BLUE}Building Docker image...${NC}"
if [ -n "$MODEL_ID" ]; then
BUILD_ARGS="$BUILD_ARGS --build-arg MODEL_ID=$MODEL_ID"
fi
docker build \
--build-arg PYTHON_VERSION=3.10 \
--build-arg VLLM_VERSION=0.6.3 \
--build-arg CUDA_VERSION=12.1.0 \
-t "${IMAGE_NAME}:${VERSION}" \
-t "${IMAGE_NAME}:latest" \
"$SCRIPT_DIR"
echo -e "${GREEN}β Docker image built successfully${NC}"
}
# Deploy locally with docker-compose
deploy_local() {
echo -e "\n${BLUE}Deploying locally with Docker Compose...${NC}"
if [ "$HAS_DOCKER" = false ]; then
echo -e "${RED}Error: Docker is required for local deployment${NC}"
exit 1
fi
# Build image
build_image
# Create .env file
ENV_FILE="${SCRIPT_DIR}/.env"
cat > "$ENV_FILE" << EOF
# Stack 2.9 Local Deployment Configuration
MODEL_ID=${MODEL_ID:-TheBloke/Llama-2-7B-Chat-AWQ}
HUGGING_FACE_TOKEN=${HF_TOKEN}
QUANTIZATION=awq
TENSOR_PARALLEL_SIZE=${TENSOR_PARALLEL_SIZE:-1}
GPU_MEMORY_UTILIZATION=${GPU_MEMORY_UTILIZATION:-0.9}
MAX_MODEL_LEN=${MAX_MODEL_LEN:-4096}
MAX_NUM_SEQS=${MAX_NUM_SEQS:-64}
MAX_NUM_BATCHED_TOKENS=${MAX_NUM_BATCHED_TOKENS:-4096}
HOST=0.0.0.0
PORT=8000
EOF
echo -e "${YELLOW}Configuration saved to $ENV_FILE${NC}"
# Start services
echo -e "\n${BLUE}Starting Stack 2.9 service...${NC}"
docker-compose -f "$COMPOSE_FILE" up -d
echo -e "\n${GREEN}β Stack 2.9 is starting...${NC}"
echo -e " API Endpoint: ${BLUE}http://localhost:8000${NC}"
echo -e " Health Check: ${BLUE}http://localhost:8000/health${NC}"
echo -e " API Docs: ${BLUE}http://localhost:8000/docs${NC}"
echo -e "\n${YELLOW}View logs: docker-compose -f '$COMPOSE_FILE' logs -f${NC}"
echo -e "${YELLOW}Stop service: docker-compose -f '$COMPOSE_FILE' down${NC}"
}
# Deploy to RunPod
deploy_runpod() {
echo -e "\n${BLUE}Deploying to RunPod...${NC}"
if ! command -v runpodctl &> /dev/null; then
echo -e "${RED}Error: runpodctl not found${NC}"
echo "Install with: npm install -g runpodctl"
exit 1
fi
# Build and push image
echo -e "${YELLOW}Building and pushing Docker image to registry...${NC}"
# User needs to set their registry
read -p "Enter your Docker registry (e.g., docker.io/username): " REGISTRY
if [ -z "$REGISTRY" ]; then
echo -e "${RED}Registry is required for RunPod deployment${NC}"
exit 1
fi
build_image
docker tag "${IMAGE_NAME}:latest" "${REGISTRY}/${IMAGE_NAME}:${VERSION}"
docker tag "${IMAGE_NAME}:latest" "${REGISTRY}/${IMAGE_NAME}:latest"
docker push "${REGISTRY}/${IMAGE_NAME}:latest"
docker push "${REGISTRY}/${IMAGE_NAME}:${VERSION}"
# Update runpod-template.json with registry
sed -i.bak "s|your-registry/stack-2.9:latest|${REGISTRY}/${IMAGE_NAME}:latest|g" \
"${SCRIPT_DIR}/runpod-template.json"
rm -f "${SCRIPT_DIR}/runpod-template.json.bak"
# Create template on RunPod
echo -e "${YELLOW}Creating RunPod template...${NC}"
runpodctl create template \
--template-file "${SCRIPT_DIR}/runpod-template.json" \
--name "stack-2.9-${VERSION}"
echo -e "\n${GREEN}β RunPod template created!${NC}"
echo -e "${YELLOW}Next steps:${NC}"
echo " 1. Go to RunPod.io and deploy using template 'stack-2.9-${VERSION}'"
echo " 2. Set your desired GPU type"
echo " 3. Configure MODEL_ID and HUGGING_FACE_TOKEN environment variables"
echo " 4. The server will start automatically on port 8000"
}
# Deploy to VastAI
deploy_vastai() {
echo -e "\n${BLUE}Deploying to Vast.ai...${NC}"
if ! command -v vastai &> /dev/null; then
echo -e "${RED}Error: vastai CLI not found${NC}"
echo "Install from: https://vast.ai/docs/cli"
exit 1
fi
# Build and push image
read -p "Enter your Docker registry (e.g., docker.io/username): " REGISTRY
if [ -z "$REGISTRY" ]; then
echo -e "${RED}Registry is required for VastAI deployment${NC}"
exit 1
fi
build_image
docker tag "${IMAGE_NAME}:latest" "${REGISTRY}/${IMAGE_NAME}:${VERSION}"
docker tag "${IMAGE_NAME}:latest" "${REGISTRY}/${IMAGE_NAME}:latest"
docker push "${REGISTRY}/${IMAGE_NAME}:latest"
docker push "${REGISTRY}/${IMAGE_NAME}:${VERSION}"
# Update vastai-template.json
sed -i.bak "s|your-registry/stack-2.9:latest|${REGISTRY}/${IMAGE_NAME}:latest|g" \
"${SCRIPT_DIR}/vastai-template.json"
rm -f "${SCRIPT_DIR}/vastai-template.json.bak"
echo -e "${GREEN}β VastAI template ready!${NC}"
echo -e "${YELLOW}Deploy with:${NC}"
echo " vastai create instance --template-file ${SCRIPT_DIR}/vastai-template.json"
echo ""
echo "Or manually:"
echo " 1. Select GPU: RTX 4090 24GB or higher"
echo " 2. Set max bid to ~\$0.50/hour"
echo " 3. Upload template and launch"
}
# Deploy to Kubernetes
deploy_kubernetes() {
echo -e "\n${BLUE}Deploying to Kubernetes...${NC}"
K8S_NAMESPACE="${K8S_NAMESPACE:-stack-2.9}"
# Check kubectl
if ! command -v kubectl &> /dev/null; then
echo -e "${RED}Error: kubectl not found${NC}"
exit 1
fi
# Build and push image (requires image builder)
read -p "Enter your Docker registry: " REGISTRY
if [ -z "$REGISTRY" ]; then
echo -e "${YELLOW}Using local image - ensure your K8s cluster can access it${NC}"
else
build_image
docker tag "${IMAGE_NAME}:latest" "${REGISTRY}/${IMAGE_NAME}:${VERSION}"
docker tag "${IMAGE_NAME}:latest" "${REGISTRY}/${IMAGE_NAME}:latest"
docker push "${REGISTRY}/${IMAGE_NAME}:latest"
IMAGE="${REGISTRY}/${IMAGE_NAME}:latest"
fi
IMAGE="${IMAGE:-${IMAGE_NAME}:latest}"
# Apply manifests
echo -e "${YELLOW}Applying Kubernetes manifests...${NC}"
# Create namespace if needed
kubectl create namespace "$K8S_NAMESPACE" --dry-run=client -o yaml | kubectl apply -f -
# Update image in deployment
sed "s|your-registry/stack-2.9:latest|${IMAGE}|g" \
"${SCRIPT_DIR}/kubernetes/deployment.yaml" | \
kubectl apply -n "$K8S_NAMESPACE" -f -
# Apply PVC
kubectl apply -n "$K8S_NAMESPACE" -f "${SCRIPT_DIR}/kubernetes/pvc.yaml"
# Apply service
kubectl apply -n "$K8S_NAMESPACE" -f "${SCRIPT_DIR}/kubernetes/service.yaml"
# Apply HPA
kubectl apply -n "$K8S_NAMESPACE" -f "${SCRIPT_DIR}/kubernetes/hpa.yaml"
echo -e "\n${GREEN}β Kubernetes deployment complete!${NC}"
echo -e " Namespace: ${K8S_NAMESPACE}"
echo -e " Checking deployment status..."
kubectl wait --for=condition=available --timeout=300s deployment/stack-2.9 -n "$K8S_NAMESPACE"
# Get service URL
SERVICE_TYPE=$(kubectl get svc stack-2.9 -n "$K8S_NAMESPACE" -o jsonpath='{.spec.type}')
if [ "$SERVICE_TYPE" = "LoadBalancer" ]; then
EXTERNAL_IP=$(kubectl get svc stack-2.9 -n "$K8S_NAMESPACE" -o jsonpath='{.status.loadBalancer.ingress[0].ip}')
echo -e " API Endpoint: ${BLUE}http://${EXTERNAL_IP}:8000${NC}"
else
NODE_PORT=$(kubectl get svc stack-2.9 -n "$K8S_NAMESPACE" -o jsonpath='{.spec.ports[0].nodePort}')
NODE_IP=$(kubectl get nodes -o jsonpath='{.items[0].status.addresses[?(@.type=="InternalIP")].address}')
echo -e " API Endpoint: ${BLUE}http://${NODE_IP}:${NODE_PORT}${NC}"
fi
}
# Health check after deployment
health_check() {
local url="http://localhost:${PORT:-8000}/health"
echo -e "\n${YELLOW}Waiting for server to become healthy...${NC}"
max_attempts=30
attempt=1
while [ $attempt -le $max_attempts ]; do
if curl -sf "$url" &> /dev/null; then
echo -e "${GREEN}β Server is healthy!${NC}"
echo -e "Health check response:"
curl -s "$url" | python3 -m json.tool 2>/dev/null || curl -s "$url"
return 0
fi
echo -n "."
sleep 2
((attempt++))
done
echo -e "\n${RED}β Health check failed after $((max_attempts * 2)) seconds${NC}"
echo "Check logs with: docker-compose -f '$COMPOSE_FILE' logs"
return 1
}
# Main execution
main() {
case "$PLATFORM" in
local|"")
detect_platform
deploy_local
health_check
;;
runpod)
detect_platform
deploy_runpod
;;
vastai)
detect_platform
deploy_vastai
;;
kubernetes|k8s)
deploy_kubernetes
;;
*)
echo -e "${RED}Unknown platform: $PLATFORM${NC}"
show_help
exit 1
;;
esac
}
# Run
parse_args "$@"
main
|