Observed KV cache size for qwen2.5:7b at fixed 16384 context varying
224MB-896MB between model loads with flash attention off. The larger
figure pushes total memory needs just past the GPU's free VRAM, so
some loads only offload 25/29 layers instead of 29/29 -- causing
DJ-agent pick latency to jump from ~1s to multiple minutes. This GPU
(6GB) has very little slack for this model/context combination even
after freeing obico and stable-diffusion's VRAM reservations.
Claude-Session: https://claude.ai/code/session_01L7Rwa6guD5wK8F8tWQwcJX
open-webui was sending OAuth credentials via both client_secret_basic
(header) and client_secret_post (body) simultaneously, which Authelia
rejects. Setting OAUTH_TOKEN_ENDPOINT_AUTH_METHOD=client_secret_post
forces a single method and matches the updated Authelia client config.
WEBUI_SECRET_KEY is now a stable env var so watchtower image updates
no longer invalidate all user sessions.
- Add open-webui OIDC client to Authelia config
- Configure open-webui with OIDC env vars pointing to Authelia
- Secret managed via AUTHELIA_OIDC_CLIENT_SECRET_OPEN_WEBUI env var in Portainer
- Add npm-network to ollama and open-webui services
- Add timezone (America/New_York) to both services
- Change open-webui image tag from main to latest
poprhythm