Files
gt-ai-os-community/docker-compose.x86.yml
HackWeasel 8b0c2b5595 fix: reduce RAM requirements for Mac/Ubuntu installers to 8GB
- Mac installer: 16GB → 8GB
- Ubuntu installer: 16GB → 8GB
- DGX installer: 64GB → 32GB (matches container reservations)
- x86 docker-compose: reduced container memory limits to match ARM64

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2025-12-29 13:16:10 -05:00

77 lines
2.0 KiB
YAML

# Docker Compose x86_64 Overlay
# Platform-specific overrides for x86_64 Linux (Ubuntu)
#
# Usage: docker compose -f docker-compose.yml -f docker-compose.x86.yml up -d
services:
# Control Panel Backend - x86 host.docker.internal
control-panel-backend:
extra_hosts:
- "host.docker.internal:host-gateway"
- "ollama-host:host-gateway"
environment:
ENVIRONMENT: production
DEBUG: "false"
# Tenant Backend - x86 host.docker.internal
tenant-backend:
extra_hosts:
- "host.docker.internal:host-gateway"
- "ollama-host:host-gateway"
environment:
ENVIRONMENT: production
DEBUG: "false"
# Resource Cluster - x86 host.docker.internal
resource-cluster:
extra_hosts:
- "host.docker.internal:host-gateway"
- "ollama-host:host-gateway"
environment:
ENVIRONMENT: production
DEBUG: "false"
# Tenant PostgreSQL Primary - x86 Performance Tuning
tenant-postgres-primary:
environment:
# x86_64 Performance settings - optimized for typical server specs
POSTGRES_SHARED_BUFFERS: 1GB
POSTGRES_EFFECTIVE_CACHE_SIZE: 3GB
POSTGRES_MAINTENANCE_WORK_MEM: 512MB
POSTGRES_MAX_CONNECTIONS: 300
POSTGRES_WORK_MEM: 128MB
deploy:
resources:
limits:
memory: 4G
reservations:
memory: 2G
labels:
- "gt2.platform=x86_64"
# VLLM Embeddings Service - x86_64 Optimized
vllm-embeddings:
platform: linux/amd64
build:
context: .
dockerfile: .deployment/docker/Dockerfile.vllm-x86
environment:
- MODEL_NAME=BAAI/bge-m3
# x86_64 optimization - adjust based on your CPU
- OMP_NUM_THREADS=8
- MKL_NUM_THREADS=8
- PYTORCH_NUM_THREADS=8
- OPENBLAS_NUM_THREADS=8
# x86_64-specific optimizations
- GT2_PLATFORM=x86_64
- MALLOC_ARENA_MAX=4
- USE_ONNX_RUNTIME=true
deploy:
resources:
limits:
memory: 4G
reservations:
memory: 3G
labels:
- "gt2.platform=x86_64"