Skip to content

Feature/docker compose optimizations #206

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
m1rl0k merged 5 commits into from
Jan 26, 2026
+14 −7
Merged

Feature/docker compose optimizations #206

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
f7755ee
feat: optimize Docker Compose with YAML anchors and configurable llam…
chris-stinemetz Jan 26, 2026
7b26d4d
fix: improve pip resilience to network timeouts in MCP Dockerfile
chris-stinemetz Jan 26, 2026
6686845
docs: add LLAMACPP_IMAGE configuration to .env.example
chris-stinemetz Jan 26, 2026
232f543
fix: correct LLAMACPP_IMAGE documentation in .env.example
chris-stinemetz Jan 26, 2026
2efe521
fix: update FakeClient.create_collection to accept on_disk_payload pa…
chris-stinemetz Jan 26, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 6 additions & 0 deletions .env.example
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -312,6 +312,12 @@ REFRAG_DECODER_MODE=prompt # prompt|soft
# Set to 0 to use Docker CPU-only server (default, stable)
USE_GPU_DECODER=0

# Llama.cpp decoder service configuration
# Default: ghcr.io/ggml-org/llama.cpp:server (multi-arch)
# CUDA support: ghcr.io/ggml-org/llama.cpp:server-cuda (for NVIDIA GPUs)
# Alternative: local builds or custom images
Copy link

@augmentcode augmentcode bot Jan 26, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The comment says server-cuda is “ARM64 specific”, but CUDA-tagged images are typically for NVIDIA/CUDA (often x86_64) and may not be ARM64/multi-arch; this could mislead users into selecting an incompatible image.

Fix This in Augment

🤖 Was this useful? React with 👍 or 👎

Copy link
Collaborator Author

@chris-stinemetz chris-stinemetz Jan 26, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Updated comment.

# LLAMACPP_IMAGE=ghcr.io/ggml-org/llama.cpp:server

REFRAG_SOFT_SCALE=1.0

# Llama.cpp runtime tuning
Expand Down
2 changes: 1 addition & 1 deletion Dockerfile.mcp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,7 +10,7 @@ ENV PYTHONDONTWRITEBYTECODE=1 \
# Python deps: reuse shared requirements file for consistency across services
# Create cache/rerank directories in same layer
COPY requirements.txt /tmp/requirements.txt
RUN pip install --no-cache-dir --upgrade -r /tmp/requirements.txt \
RUN pip install --no-cache-dir --upgrade --timeout 300 --retries 3 -r /tmp/requirements.txt \
&& mkdir -p /tmp/cache && chmod 755 /tmp/cache \
&& mkdir -p /tmp/rerank_events /tmp/rerank_weights \
&& chmod 777 /tmp/rerank_events /tmp/rerank_weights
Expand Down
10 changes: 5 additions & 5 deletions docker-compose.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -66,11 +66,11 @@ services:
container_name: redis-cache
ports:
- "6379:6379"
command: ["redis-server", "--appendonly", "yes", "--maxmemory", "256mb", "--maxmemory-policy", "allkeys-lru"]
command: [ "redis-server", "--appendonly", "yes", "--maxmemory", "256mb", "--maxmemory-policy", "allkeys-lru" ]
volumes:
- redis_data:/data
healthcheck:
test: ["CMD", "redis-cli", "ping"]
test: [ "CMD", "redis-cli", "ping" ]
interval: 10s
timeout: 5s
retries: 5
Expand Down Expand Up @@ -436,7 +436,7 @@ services:

# Llama.cpp decoder service - same as base compose
llamacpp:
image: ghcr.io/ggml-org/llama.cpp:server
image: ${LLAMACPP_IMAGE:-ghcr.io/ggml-org/llama.cpp:server}
container_name: llama-decoder-dev-remote
environment:
- LLAMA_ARG_MODEL=/models/model.gguf
Expand Down Expand Up @@ -518,8 +518,8 @@ services:
volumes:
- workspace_pvc:/work:rw
- codebase_pvc:/work/.codebase:rw
entrypoint: ["sh", "-c", "mkdir -p /tmp/logs /tmp/huggingface/hub /tmp/huggingface/transformers /tmp/huggingface/fastembed && /app/scripts/wait-for-qdrant.sh && cd /app && python /app/scripts/ingest_code.py --root /work"]
restart: "no" # Run once on startup, do not restart after completion
entrypoint: [ "sh", "-c", "mkdir -p /tmp/logs /tmp/huggingface/hub /tmp/huggingface/transformers /tmp/huggingface/fastembed && /app/scripts/wait-for-qdrant.sh && cd /app && python /app/scripts/ingest_code.py --root /work" ]
restart: "no" # Run once on startup, do not restart after completion
cpus: 4.0
networks:
- dev-remote-network
Expand Down
3 changes: 2 additions & 1 deletion tests/test_ingest_schema_mode.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,14 +30,15 @@ def get_collection(self, name):
payload_schema=self.payload_schema,
)

def create_collection(self, collection_name, vectors_config, sparse_vectors_config=None, hnsw_config=None, quantization_config=None):
def create_collection(self, collection_name, vectors_config, sparse_vectors_config=None, hnsw_config=None, quantization_config=None, on_disk_payload=None):
self.create_calls.append(
{
"collection_name": collection_name,
"vectors_config": vectors_config,
"sparse_vectors_config": sparse_vectors_config,
"hnsw_config": hnsw_config,
"quantization_config": quantization_config,
"on_disk_payload": on_disk_payload,
}
)
self.collection_exists = True
Expand Down