dschlueter/llamacppctl
1
0
Fork 0
Code Issues Pull requests Projects Releases 1 Packages Wiki Activity

build(docker): pin the llama.cpp image by digest

Browse source 
The :server-cuda tag is a moving target, so a fresh pull could silently
change server behaviour (flags, the baked-in healthcheck, ...). Pin the
default image to the current digest for reproducibility; overriding `image`
in the config or via --image still works. A comment documents how to bump
the pin.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
Dieter Schlüter 2026-07-07 10:28:23 +02:00
commit c553654aae
3 changed files with 9 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

3
docs/SECURITY_AND_OPERATIONS.md
View file

@ -34,7 +34,8 @@ testbar von der Orchestrierung.
1. **Eingebaute Defaults** (`config.builtin_defaults()`) — spiegeln die 1. **Eingebaute Defaults** (`config.builtin_defaults()`) — spiegeln die
Standardwerte der ursprünglichen Shell-Skripte wider (Image Standardwerte der ursprünglichen Shell-Skripte wider (Image
`ghcr.io/ggml-org/llama.cpp:server-cuda`, `host_port=8001`, `ghcr.io/ggml-org/llama.cpp` per Digest gepinnt für Reproduzierbarkeit,
`host_port=8001`,
`container_name=va_llm`, `gpu_device=1`, `jinja/fa/kv_unified/ `container_name=va_llm`, `gpu_device=1`, `jinja/fa/kv_unified/
cont_batching/no_context_shift=true`, `reasoning=on`, cont_batching/no_context_shift=true`, `reasoning=on`,
`cache_type_k/v=q4_0`, `batch_size=1024`, `ubatch_size=512`, `cache_type_k/v=q4_0`, `batch_size=1024`, `ubatch_size=512`,

4
llama.cpp.config.example
View file

@ -14,7 +14,9 @@
# locking (--change), and --stop/--check targeting. # locking (--change), and --stop/--check targeting.
[default] [default]
image = ghcr.io/ggml-org/llama.cpp:server-cuda # Pinned by digest for reproducibility. The :server-cuda tag is a moving target;
# to update, pull it, read the new digest, and replace the pin below.
image = ghcr.io/ggml-org/llama.cpp@sha256:5535de118ed457f761cbfeacd7e10fef31cb391ca7cac1d5c78b11d28fcf88e6
# hf_home unterstützt Environment-Variablen und ~, z. B. hf_home = ${HF_HOME} # hf_home unterstützt Environment-Variablen und ~, z. B. hf_home = ${HF_HOME}
hf_home = /srv/models hf_home = /srv/models
model_path = qwen3/default.gguf model_path = qwen3/default.gguf

5
src/llamacppctl/config.py
View file

@ -36,7 +36,10 @@ class ConfigError(ValueError):
def builtin_defaults() -> dict: def builtin_defaults() -> dict:
"""Mirrors the original start-llm-server.sh / status-llm-server.sh defaults.""" """Mirrors the original start-llm-server.sh / status-llm-server.sh defaults."""
return { return {
"image": "ghcr.io/ggml-org/llama.cpp:server-cuda", # Pinned by digest for reproducibility (the :server-cuda tag is a moving
# target). To update: docker pull ghcr.io/ggml-org/llama.cpp:server-cuda,
# read the new digest, and bump it here (or override `image` in the config).
"image": "ghcr.io/ggml-org/llama.cpp@sha256:5535de118ed457f761cbfeacd7e10fef31cb391ca7cac1d5c78b11d28fcf88e6",
"hf_home": "/models", "hf_home": "/models",
"model_path": "qwen3/default.gguf", "model_path": "qwen3/default.gguf",
"container_name": "va_llm", "container_name": "va_llm",