homelab-codex-ws/services/kb-query/app/main.py

135 lines
6.2 KiB
Python
Raw Normal View History

"""kb-query -- module 5 phase 4 (kb/phases/kb-m5-faza4.md §4): first user-facing HTTP
feat(kb): aktywny fallback embeddingów SOLARIA→PIHA dla kb-query (faza 4 Krok 2) Ostatni krok fazy 4 KB (plan §2 Decyzja 2, §5): kb-query przestaje być martwe przez ~16 h/dobę, gdy SOLARIA (GPU) śpi — zapytania embeduje wtedy lokalna Ollama CPU na PIHA (wolniej: ~790 ms+ vs ~207 ms na GPU, ale działa). Nowy serwis services/ollama-piha (GitOps, owner_node: piha): - ollama/ollama:latest (arm64 natywnie), OLLAMA_KEEP_ALIVE=0 — model zwalnia RAM natychmiast po każdym wywołaniu (spike, nie rezydent; PIHA dzieli 8 GB z HA) - bind wyłącznie 127.0.0.1 + LAN_BIND_IP (192.168.31.5), nigdy 0.0.0.0/Tailscale - named volume ollama_piha_models (NVMe data-root) zamiast bind-mounta — obraz biega jako root w kontenerze i bind łamałby wzorzec uid PIHA (oskar=1004, kontenery uid 1000, setgid pi) - override hosts/piha/runtime/ollama-piha: mem_limit 2560m (wartość startowa z planu, do potwierdzenia kalibracją na żywo), świadomie bez mem_reservation - pull bge-m3 to jawny, ręczny krok deployu (README) — obraz nie ma modeli kb-query — maszyna stanów fallbacku (app/embed_router.py): - health-check SOLARII (GET /api/tags, timeout 1.5 s) z cache 30 s — zero sondowania per request; po powrocie SOLARII ruch wraca na GPU w ≤30 s - primary up → embed na SOLARII z twardym timeoutem 3 s; błąd W TRAKCIE zapytania = jednorazowe przełączenie (krok 3b planu): status down na 30 s i TO SAMO zapytanie leci na fallback — user nie widzi błędu SOLARII - primary down → embed prosto na ollama-piha (bez twardego timeoutu: CPU + zimny load modelu to legalnie pojedyncze sekundy) - 503 tylko gdy oba backendy padłe (lub fallback nieskonfigurowany) - inwariant modelu, druga połowa: każdy backend weryfikowany raz, leniwie przy pierwszym użyciu, że /api/tags zawiera EMBED_MODEL (bge-m3 — ta sama wartość co startowy check przeciw document_chunk.model/document_summary.embedding_model); niezgodność = ERROR log + 500, nigdy ciche liczenie dystansów między różnymi przestrzeniami embeddingów; leniwie, bo śpiąca SOLARIA nie może blokować startu serwisu - odpowiedź /search: nowe pole embed_backend ("solaria"|"piha") + sol_status wg realnego świata routera (UI już renderuje down jako "offline (fallback embed)"); log INFO backend=... elapsed_ms=... per zapytanie - /healthz: sol_status przez cache routera (spójny widok z routingiem) + fallback_status (żywa, tania sonda /api/tags) Konfiguracja spójnie przez env (compose + env.example + service.yaml + README): EMBED_PRIMARY_URL (zastępuje OLLAMA_URL), EMBED_FALLBACK_URL (pusty = brak fallbacku, zachowanie sprzed kroku 2), EMBED_{PRIMARY,FALLBACK}_NAME, EMBED_HEALTH_TTL_S/EMBED_HEALTH_TIMEOUT_S/EMBED_PRIMARY_TIMEOUT_S. Testy: 39 pass (14 nowych w test_embed_router.py: cache TTL, failover w trakcie zapytania, powrót po TTL, oba padłe, mismatch modelu na primary i fallbacku, tag "bge-m3:latest" vs "bge-m3"); docker build + smoke (importy + uvicorn do guardu KB_DSN) OK; compose config OK dla obu stacków. Deploy (Oskar, na PIHA z mastera po merge): cd ~/homelab-codex-ws && git pull # 1. ollama-piha cp services/ollama-piha/env.example services/ollama-piha/.env docker compose -f services/ollama-piha/docker-compose.yml \ -f hosts/piha/runtime/ollama-piha/docker-compose.override.yml \ --env-file services/ollama-piha/.env up -d docker exec ollama-piha ollama pull bge-m3 # ręczny krok, obowiązkowy services/ollama-piha/healthcheck.sh # 2. kb-query (dopisać fallback do istniejącego .env) echo 'EMBED_FALLBACK_URL=http://192.168.31.5:11434' >> services/kb-query/.env docker compose -f services/kb-query/docker-compose.yml \ -f hosts/piha/runtime/kb-query/docker-compose.override.yml up -d --build services/kb-query/healthcheck.sh # (deploy-node.sh też podniesie oba serwisy z hosts/piha/services.yaml, # ale pull bge-m3 i .env pozostają ręczne) Weryfikacja: testy A/B/C w services/kb-query/README.md (backend=solaria przy SOLARII online; backend=piha przy symulacji offline; powrót na GPU w ≤30 s). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-29 19:01:29 +02:00
entry point to the KB. Wraps `kb_retrieval` retrieval (module 5 phase 3, gated PASS --
docs/sessions/2026-07-21.md) in FastAPI. This is a search API, not chat: no answer synthesis,
no LLM call over the results (that is phase 5, out of scope here).
Embed path (Krok 2, plan §2 decision 2 / §5 -- the active fallback): queries embed through
`app.embed_router.EmbedRouter` -- SOLARIA's GPU Ollama (EMBED_PRIMARY_URL) when its cached
~30 s health-check says up, the local CPU `ollama-piha` (EMBED_FALLBACK_URL) when SOLARIA
sleeps. A mid-query primary failure fails over the SAME request instead of surfacing an
error. Only when BOTH backends are unavailable does `/search` return 503; a backend serving
the wrong model (vs the bge-m3 space `document_chunk` is indexed in) is a loud 500, never a
silent cross-space distance computation.
feat(kb-query): add search frontend (module 5 phase 4, plan §7, Krok 4) Krok 4 of the phase-4 plan done ahead of the local-embed-fallback step (Krok 2, deliberately deferred -- embed stays a plain SOLARIA call, per task instruction): one FastAPI process now serves both the /search API and the UI, no separate frontend build (plan §2 decision 4). - GET / renders a Jinja2 shell; app/static/app.js (vanilla, no build) and style.css are the whole client. Query -> /search, results grouped by envelope_id client-side (chunks sorted by dist, <details> fragments). - Colour thresholds per plan §7: dist<0.45 green, 0.45-0.55 yellow (still shown with a warning), >0.55 never rendered as an individual result; if a query ends up with nothing renderable, one "Brak odpowiedzi w KB" message replaces the list, carrying the best observed dist. - Paperless hits link out; gmail hits get a "kopiuj Message-ID" button (there's nothing to link to yet, plan §2 decision 3) plus header metadata. Cascade/flat toggle defaults to cascade. Footer shows sol_status, refreshed from /healthz on load and after each search. - /search gained additive summary/summary_tags fields (document_summary, haiku track) so the UI can show a document summary as each result group's header -- non-breaking, existing response shape untouched. - Tests: app/db.py + app/search.py unit tests (mocked DB/HTTP, no live deps) cover the new summary join; tests/test_frontend.py drives GET / and /static/* via TestClient without running the DB-requiring lifespan; tests/frontend/app.test.js (Node's built-in test runner, no framework) covers query-URL encoding, threshold colouring, and envelope grouping. - Verified live: docker build + container against kb-postgres@PIHA over LAN and Ollama@SOLARIA over Tailscale -- GET / (HTML), /static/app.js, /healthz, and /search (cascade + flat) all round-tripped correctly, including real summary/summary_tags data. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:39:46 +02:00
`GET /` (Krok 4, plan §7) serves the search UI from this same FastAPI process -- one image, one
container (plan §2 decision 4): a Jinja2 shell + a static vanilla-JS file, no node build step.
`/` and `/static/*` need no DB/Ollama, so they stay reachable even while `/search` is 503ing.
feat(kb-query): domyslny mode /search = hybrid (DoD (d) fazy mailowej) Eval na pelnym korpusie 2026-08-06 (187 025 zembedowanych chunkow mailowych w HNSW) dal PASS: kryterium 1 (regresja paperless) bez degradacji zadnego istniejacego hitu we flat ani w hybrid, mailowe hit@3 5/5. Koszt hybrydy to jedno dodatkowe zapytanie SQL na wyszukiwanie. Surowe wyniki: eval-http-2026-08-06.json / eval-direct-2026-08-06.json w ~/kb/mail/ingest-logs na PIHA (niecommitowane, artefakt runu). - app/main.py: Query("cascade") -> Query("hybrid"); pattern bez zmian, wiec jawne ?mode=cascade i ?mode=flat dzialaja dokladnie jak dotad. - app/static/app.js: przy odznaczonym "tryb flat (debug)" UI nie wysyla juz parametru mode w ogole -- dziedziczy default API. Default zdefiniowany w jednym miejscu (serwer), nie zduplikowany w JS. - testy: nowa klasa TestSearchEndpointModeDefault (TestClient bez lifespan, fake pool/router) sprawdza kontrakt HTTP -- brak mode => tor hybrid (weryfikowany po obecnosci koperty gmail osiagalnej wylacznie galezia hybrid, nie po samej etykiecie), jawne mode=flat / mode=cascade => stare tory, nieznany mode => 422. Frontend: buildSearchUrl pomija mode gdy brak. - docs: kb/services/kb-query.md (tabela trybow + endpoint + przyklad odpowiedzi + opis przelacznika w UI), env.example/service.yaml (komentarze SUMMARY_MODEL; default mode nie jest konfigurowalny przez env), kb/phases/kb-m5-faza-mailowa.md (DoD (d) SPELNIONE 2026-08-06 + wzmianki w Kroku 3, Wyniku bramki, decyzjach Etapu B i tabeli planu). Weryfikacja: pytest services/kb-query -> 46 passed; node --test tests/frontend/app.test.js -> 6/6; docker build OK + smoke run (uvicorn startuje, bez KB_DSN swiadomie konczy sie RuntimeError z env.example). Deploy NIE wykonany -- operator wdraza z mastera na PIHA po mergu. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-06 13:30:02 +02:00
`mode=hybrid` (faza mailowa, plan Krok 3, kb/phases/kb-m5-faza-mailowa.md §6) is the
DEFAULT since 2026-08-06 (DoD (d) of that phase): the quality gate (§8) re-ran on the full
mail corpus (187k embedded mail chunks in HNSW) and PASSed -- no paperless hit degraded,
mail hit@3 5/5 -- so mail bodies are now visible without asking for them. Cost: one extra
SQL query per search (the summaryless-source chunk scan). `mode=cascade` (documents only,
the old default) and `mode=flat` (debug, no summary pre-filter) stay available explicitly
and behave exactly as before.
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
"""
from __future__ import annotations
feat(kb): aktywny fallback embeddingów SOLARIA→PIHA dla kb-query (faza 4 Krok 2) Ostatni krok fazy 4 KB (plan §2 Decyzja 2, §5): kb-query przestaje być martwe przez ~16 h/dobę, gdy SOLARIA (GPU) śpi — zapytania embeduje wtedy lokalna Ollama CPU na PIHA (wolniej: ~790 ms+ vs ~207 ms na GPU, ale działa). Nowy serwis services/ollama-piha (GitOps, owner_node: piha): - ollama/ollama:latest (arm64 natywnie), OLLAMA_KEEP_ALIVE=0 — model zwalnia RAM natychmiast po każdym wywołaniu (spike, nie rezydent; PIHA dzieli 8 GB z HA) - bind wyłącznie 127.0.0.1 + LAN_BIND_IP (192.168.31.5), nigdy 0.0.0.0/Tailscale - named volume ollama_piha_models (NVMe data-root) zamiast bind-mounta — obraz biega jako root w kontenerze i bind łamałby wzorzec uid PIHA (oskar=1004, kontenery uid 1000, setgid pi) - override hosts/piha/runtime/ollama-piha: mem_limit 2560m (wartość startowa z planu, do potwierdzenia kalibracją na żywo), świadomie bez mem_reservation - pull bge-m3 to jawny, ręczny krok deployu (README) — obraz nie ma modeli kb-query — maszyna stanów fallbacku (app/embed_router.py): - health-check SOLARII (GET /api/tags, timeout 1.5 s) z cache 30 s — zero sondowania per request; po powrocie SOLARII ruch wraca na GPU w ≤30 s - primary up → embed na SOLARII z twardym timeoutem 3 s; błąd W TRAKCIE zapytania = jednorazowe przełączenie (krok 3b planu): status down na 30 s i TO SAMO zapytanie leci na fallback — user nie widzi błędu SOLARII - primary down → embed prosto na ollama-piha (bez twardego timeoutu: CPU + zimny load modelu to legalnie pojedyncze sekundy) - 503 tylko gdy oba backendy padłe (lub fallback nieskonfigurowany) - inwariant modelu, druga połowa: każdy backend weryfikowany raz, leniwie przy pierwszym użyciu, że /api/tags zawiera EMBED_MODEL (bge-m3 — ta sama wartość co startowy check przeciw document_chunk.model/document_summary.embedding_model); niezgodność = ERROR log + 500, nigdy ciche liczenie dystansów między różnymi przestrzeniami embeddingów; leniwie, bo śpiąca SOLARIA nie może blokować startu serwisu - odpowiedź /search: nowe pole embed_backend ("solaria"|"piha") + sol_status wg realnego świata routera (UI już renderuje down jako "offline (fallback embed)"); log INFO backend=... elapsed_ms=... per zapytanie - /healthz: sol_status przez cache routera (spójny widok z routingiem) + fallback_status (żywa, tania sonda /api/tags) Konfiguracja spójnie przez env (compose + env.example + service.yaml + README): EMBED_PRIMARY_URL (zastępuje OLLAMA_URL), EMBED_FALLBACK_URL (pusty = brak fallbacku, zachowanie sprzed kroku 2), EMBED_{PRIMARY,FALLBACK}_NAME, EMBED_HEALTH_TTL_S/EMBED_HEALTH_TIMEOUT_S/EMBED_PRIMARY_TIMEOUT_S. Testy: 39 pass (14 nowych w test_embed_router.py: cache TTL, failover w trakcie zapytania, powrót po TTL, oba padłe, mismatch modelu na primary i fallbacku, tag "bge-m3:latest" vs "bge-m3"); docker build + smoke (importy + uvicorn do guardu KB_DSN) OK; compose config OK dla obu stacków. Deploy (Oskar, na PIHA z mastera po merge): cd ~/homelab-codex-ws && git pull # 1. ollama-piha cp services/ollama-piha/env.example services/ollama-piha/.env docker compose -f services/ollama-piha/docker-compose.yml \ -f hosts/piha/runtime/ollama-piha/docker-compose.override.yml \ --env-file services/ollama-piha/.env up -d docker exec ollama-piha ollama pull bge-m3 # ręczny krok, obowiązkowy services/ollama-piha/healthcheck.sh # 2. kb-query (dopisać fallback do istniejącego .env) echo 'EMBED_FALLBACK_URL=http://192.168.31.5:11434' >> services/kb-query/.env docker compose -f services/kb-query/docker-compose.yml \ -f hosts/piha/runtime/kb-query/docker-compose.override.yml up -d --build services/kb-query/healthcheck.sh # (deploy-node.sh też podniesie oba serwisy z hosts/piha/services.yaml, # ale pull bge-m3 i .env pozostają ręczne) Weryfikacja: testy A/B/C w services/kb-query/README.md (backend=solaria przy SOLARII online; backend=piha przy symulacji offline; powrót na GPU w ≤30 s). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-29 19:01:29 +02:00
import logging
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
import os
feat(kb-query): add search frontend (module 5 phase 4, plan §7, Krok 4) Krok 4 of the phase-4 plan done ahead of the local-embed-fallback step (Krok 2, deliberately deferred -- embed stays a plain SOLARIA call, per task instruction): one FastAPI process now serves both the /search API and the UI, no separate frontend build (plan §2 decision 4). - GET / renders a Jinja2 shell; app/static/app.js (vanilla, no build) and style.css are the whole client. Query -> /search, results grouped by envelope_id client-side (chunks sorted by dist, <details> fragments). - Colour thresholds per plan §7: dist<0.45 green, 0.45-0.55 yellow (still shown with a warning), >0.55 never rendered as an individual result; if a query ends up with nothing renderable, one "Brak odpowiedzi w KB" message replaces the list, carrying the best observed dist. - Paperless hits link out; gmail hits get a "kopiuj Message-ID" button (there's nothing to link to yet, plan §2 decision 3) plus header metadata. Cascade/flat toggle defaults to cascade. Footer shows sol_status, refreshed from /healthz on load and after each search. - /search gained additive summary/summary_tags fields (document_summary, haiku track) so the UI can show a document summary as each result group's header -- non-breaking, existing response shape untouched. - Tests: app/db.py + app/search.py unit tests (mocked DB/HTTP, no live deps) cover the new summary join; tests/test_frontend.py drives GET / and /static/* via TestClient without running the DB-requiring lifespan; tests/frontend/app.test.js (Node's built-in test runner, no framework) covers query-URL encoding, threshold colouring, and envelope grouping. - Verified live: docker build + container against kb-postgres@PIHA over LAN and Ollama@SOLARIA over Tailscale -- GET / (HTML), /static/app.js, /healthz, and /search (cascade + flat) all round-tripped correctly, including real summary/summary_tags data. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:39:46 +02:00
import pathlib
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
from contextlib import asynccontextmanager
import aiohttp
feat(kb-query): add search frontend (module 5 phase 4, plan §7, Krok 4) Krok 4 of the phase-4 plan done ahead of the local-embed-fallback step (Krok 2, deliberately deferred -- embed stays a plain SOLARIA call, per task instruction): one FastAPI process now serves both the /search API and the UI, no separate frontend build (plan §2 decision 4). - GET / renders a Jinja2 shell; app/static/app.js (vanilla, no build) and style.css are the whole client. Query -> /search, results grouped by envelope_id client-side (chunks sorted by dist, <details> fragments). - Colour thresholds per plan §7: dist<0.45 green, 0.45-0.55 yellow (still shown with a warning), >0.55 never rendered as an individual result; if a query ends up with nothing renderable, one "Brak odpowiedzi w KB" message replaces the list, carrying the best observed dist. - Paperless hits link out; gmail hits get a "kopiuj Message-ID" button (there's nothing to link to yet, plan §2 decision 3) plus header metadata. Cascade/flat toggle defaults to cascade. Footer shows sol_status, refreshed from /healthz on load and after each search. - /search gained additive summary/summary_tags fields (document_summary, haiku track) so the UI can show a document summary as each result group's header -- non-breaking, existing response shape untouched. - Tests: app/db.py + app/search.py unit tests (mocked DB/HTTP, no live deps) cover the new summary join; tests/test_frontend.py drives GET / and /static/* via TestClient without running the DB-requiring lifespan; tests/frontend/app.test.js (Node's built-in test runner, no framework) covers query-URL encoding, threshold colouring, and envelope grouping. - Verified live: docker build + container against kb-postgres@PIHA over LAN and Ollama@SOLARIA over Tailscale -- GET / (HTML), /static/app.js, /healthz, and /search (cascade + flat) all round-tripped correctly, including real summary/summary_tags data. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:39:46 +02:00
from fastapi import FastAPI, HTTPException, Query, Request
from fastapi.responses import HTMLResponse
from fastapi.staticfiles import StaticFiles
from fastapi.templating import Jinja2Templates
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
from app.db import create_pool
feat(kb): aktywny fallback embeddingów SOLARIA→PIHA dla kb-query (faza 4 Krok 2) Ostatni krok fazy 4 KB (plan §2 Decyzja 2, §5): kb-query przestaje być martwe przez ~16 h/dobę, gdy SOLARIA (GPU) śpi — zapytania embeduje wtedy lokalna Ollama CPU na PIHA (wolniej: ~790 ms+ vs ~207 ms na GPU, ale działa). Nowy serwis services/ollama-piha (GitOps, owner_node: piha): - ollama/ollama:latest (arm64 natywnie), OLLAMA_KEEP_ALIVE=0 — model zwalnia RAM natychmiast po każdym wywołaniu (spike, nie rezydent; PIHA dzieli 8 GB z HA) - bind wyłącznie 127.0.0.1 + LAN_BIND_IP (192.168.31.5), nigdy 0.0.0.0/Tailscale - named volume ollama_piha_models (NVMe data-root) zamiast bind-mounta — obraz biega jako root w kontenerze i bind łamałby wzorzec uid PIHA (oskar=1004, kontenery uid 1000, setgid pi) - override hosts/piha/runtime/ollama-piha: mem_limit 2560m (wartość startowa z planu, do potwierdzenia kalibracją na żywo), świadomie bez mem_reservation - pull bge-m3 to jawny, ręczny krok deployu (README) — obraz nie ma modeli kb-query — maszyna stanów fallbacku (app/embed_router.py): - health-check SOLARII (GET /api/tags, timeout 1.5 s) z cache 30 s — zero sondowania per request; po powrocie SOLARII ruch wraca na GPU w ≤30 s - primary up → embed na SOLARII z twardym timeoutem 3 s; błąd W TRAKCIE zapytania = jednorazowe przełączenie (krok 3b planu): status down na 30 s i TO SAMO zapytanie leci na fallback — user nie widzi błędu SOLARII - primary down → embed prosto na ollama-piha (bez twardego timeoutu: CPU + zimny load modelu to legalnie pojedyncze sekundy) - 503 tylko gdy oba backendy padłe (lub fallback nieskonfigurowany) - inwariant modelu, druga połowa: każdy backend weryfikowany raz, leniwie przy pierwszym użyciu, że /api/tags zawiera EMBED_MODEL (bge-m3 — ta sama wartość co startowy check przeciw document_chunk.model/document_summary.embedding_model); niezgodność = ERROR log + 500, nigdy ciche liczenie dystansów między różnymi przestrzeniami embeddingów; leniwie, bo śpiąca SOLARIA nie może blokować startu serwisu - odpowiedź /search: nowe pole embed_backend ("solaria"|"piha") + sol_status wg realnego świata routera (UI już renderuje down jako "offline (fallback embed)"); log INFO backend=... elapsed_ms=... per zapytanie - /healthz: sol_status przez cache routera (spójny widok z routingiem) + fallback_status (żywa, tania sonda /api/tags) Konfiguracja spójnie przez env (compose + env.example + service.yaml + README): EMBED_PRIMARY_URL (zastępuje OLLAMA_URL), EMBED_FALLBACK_URL (pusty = brak fallbacku, zachowanie sprzed kroku 2), EMBED_{PRIMARY,FALLBACK}_NAME, EMBED_HEALTH_TTL_S/EMBED_HEALTH_TIMEOUT_S/EMBED_PRIMARY_TIMEOUT_S. Testy: 39 pass (14 nowych w test_embed_router.py: cache TTL, failover w trakcie zapytania, powrót po TTL, oba padłe, mismatch modelu na primary i fallbacku, tag "bge-m3:latest" vs "bge-m3"); docker build + smoke (importy + uvicorn do guardu KB_DSN) OK; compose config OK dla obu stacków. Deploy (Oskar, na PIHA z mastera po merge): cd ~/homelab-codex-ws && git pull # 1. ollama-piha cp services/ollama-piha/env.example services/ollama-piha/.env docker compose -f services/ollama-piha/docker-compose.yml \ -f hosts/piha/runtime/ollama-piha/docker-compose.override.yml \ --env-file services/ollama-piha/.env up -d docker exec ollama-piha ollama pull bge-m3 # ręczny krok, obowiązkowy services/ollama-piha/healthcheck.sh # 2. kb-query (dopisać fallback do istniejącego .env) echo 'EMBED_FALLBACK_URL=http://192.168.31.5:11434' >> services/kb-query/.env docker compose -f services/kb-query/docker-compose.yml \ -f hosts/piha/runtime/kb-query/docker-compose.override.yml up -d --build services/kb-query/healthcheck.sh # (deploy-node.sh też podniesie oba serwisy z hosts/piha/services.yaml, # ale pull bge-m3 i .env pozostają ręczne) Weryfikacja: testy A/B/C w services/kb-query/README.md (backend=solaria przy SOLARII online; backend=piha przy symulacji offline; powrót na GPU w ≤30 s). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-29 19:01:29 +02:00
from app.embed_router import EmbedBackendError, EmbedRouter, ModelMismatchError
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
from app.search import run_search
from app.startup import validate_embed_model
feat(kb): aktywny fallback embeddingów SOLARIA→PIHA dla kb-query (faza 4 Krok 2) Ostatni krok fazy 4 KB (plan §2 Decyzja 2, §5): kb-query przestaje być martwe przez ~16 h/dobę, gdy SOLARIA (GPU) śpi — zapytania embeduje wtedy lokalna Ollama CPU na PIHA (wolniej: ~790 ms+ vs ~207 ms na GPU, ale działa). Nowy serwis services/ollama-piha (GitOps, owner_node: piha): - ollama/ollama:latest (arm64 natywnie), OLLAMA_KEEP_ALIVE=0 — model zwalnia RAM natychmiast po każdym wywołaniu (spike, nie rezydent; PIHA dzieli 8 GB z HA) - bind wyłącznie 127.0.0.1 + LAN_BIND_IP (192.168.31.5), nigdy 0.0.0.0/Tailscale - named volume ollama_piha_models (NVMe data-root) zamiast bind-mounta — obraz biega jako root w kontenerze i bind łamałby wzorzec uid PIHA (oskar=1004, kontenery uid 1000, setgid pi) - override hosts/piha/runtime/ollama-piha: mem_limit 2560m (wartość startowa z planu, do potwierdzenia kalibracją na żywo), świadomie bez mem_reservation - pull bge-m3 to jawny, ręczny krok deployu (README) — obraz nie ma modeli kb-query — maszyna stanów fallbacku (app/embed_router.py): - health-check SOLARII (GET /api/tags, timeout 1.5 s) z cache 30 s — zero sondowania per request; po powrocie SOLARII ruch wraca na GPU w ≤30 s - primary up → embed na SOLARII z twardym timeoutem 3 s; błąd W TRAKCIE zapytania = jednorazowe przełączenie (krok 3b planu): status down na 30 s i TO SAMO zapytanie leci na fallback — user nie widzi błędu SOLARII - primary down → embed prosto na ollama-piha (bez twardego timeoutu: CPU + zimny load modelu to legalnie pojedyncze sekundy) - 503 tylko gdy oba backendy padłe (lub fallback nieskonfigurowany) - inwariant modelu, druga połowa: każdy backend weryfikowany raz, leniwie przy pierwszym użyciu, że /api/tags zawiera EMBED_MODEL (bge-m3 — ta sama wartość co startowy check przeciw document_chunk.model/document_summary.embedding_model); niezgodność = ERROR log + 500, nigdy ciche liczenie dystansów między różnymi przestrzeniami embeddingów; leniwie, bo śpiąca SOLARIA nie może blokować startu serwisu - odpowiedź /search: nowe pole embed_backend ("solaria"|"piha") + sol_status wg realnego świata routera (UI już renderuje down jako "offline (fallback embed)"); log INFO backend=... elapsed_ms=... per zapytanie - /healthz: sol_status przez cache routera (spójny widok z routingiem) + fallback_status (żywa, tania sonda /api/tags) Konfiguracja spójnie przez env (compose + env.example + service.yaml + README): EMBED_PRIMARY_URL (zastępuje OLLAMA_URL), EMBED_FALLBACK_URL (pusty = brak fallbacku, zachowanie sprzed kroku 2), EMBED_{PRIMARY,FALLBACK}_NAME, EMBED_HEALTH_TTL_S/EMBED_HEALTH_TIMEOUT_S/EMBED_PRIMARY_TIMEOUT_S. Testy: 39 pass (14 nowych w test_embed_router.py: cache TTL, failover w trakcie zapytania, powrót po TTL, oba padłe, mismatch modelu na primary i fallbacku, tag "bge-m3:latest" vs "bge-m3"); docker build + smoke (importy + uvicorn do guardu KB_DSN) OK; compose config OK dla obu stacków. Deploy (Oskar, na PIHA z mastera po merge): cd ~/homelab-codex-ws && git pull # 1. ollama-piha cp services/ollama-piha/env.example services/ollama-piha/.env docker compose -f services/ollama-piha/docker-compose.yml \ -f hosts/piha/runtime/ollama-piha/docker-compose.override.yml \ --env-file services/ollama-piha/.env up -d docker exec ollama-piha ollama pull bge-m3 # ręczny krok, obowiązkowy services/ollama-piha/healthcheck.sh # 2. kb-query (dopisać fallback do istniejącego .env) echo 'EMBED_FALLBACK_URL=http://192.168.31.5:11434' >> services/kb-query/.env docker compose -f services/kb-query/docker-compose.yml \ -f hosts/piha/runtime/kb-query/docker-compose.override.yml up -d --build services/kb-query/healthcheck.sh # (deploy-node.sh też podniesie oba serwisy z hosts/piha/services.yaml, # ale pull bge-m3 i .env pozostają ręczne) Weryfikacja: testy A/B/C w services/kb-query/README.md (backend=solaria przy SOLARII online; backend=piha przy symulacji offline; powrót na GPU w ≤30 s). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-29 19:01:29 +02:00
# uvicorn only configures its own loggers; without this the router's
# backend=solaria/piha lines (kb-query.embed) would never reach docker logs.
logging.basicConfig(level=logging.INFO, format="%(asctime)s %(levelname)s %(name)s %(message)s")
feat(kb-query): add search frontend (module 5 phase 4, plan §7, Krok 4) Krok 4 of the phase-4 plan done ahead of the local-embed-fallback step (Krok 2, deliberately deferred -- embed stays a plain SOLARIA call, per task instruction): one FastAPI process now serves both the /search API and the UI, no separate frontend build (plan §2 decision 4). - GET / renders a Jinja2 shell; app/static/app.js (vanilla, no build) and style.css are the whole client. Query -> /search, results grouped by envelope_id client-side (chunks sorted by dist, <details> fragments). - Colour thresholds per plan §7: dist<0.45 green, 0.45-0.55 yellow (still shown with a warning), >0.55 never rendered as an individual result; if a query ends up with nothing renderable, one "Brak odpowiedzi w KB" message replaces the list, carrying the best observed dist. - Paperless hits link out; gmail hits get a "kopiuj Message-ID" button (there's nothing to link to yet, plan §2 decision 3) plus header metadata. Cascade/flat toggle defaults to cascade. Footer shows sol_status, refreshed from /healthz on load and after each search. - /search gained additive summary/summary_tags fields (document_summary, haiku track) so the UI can show a document summary as each result group's header -- non-breaking, existing response shape untouched. - Tests: app/db.py + app/search.py unit tests (mocked DB/HTTP, no live deps) cover the new summary join; tests/test_frontend.py drives GET / and /static/* via TestClient without running the DB-requiring lifespan; tests/frontend/app.test.js (Node's built-in test runner, no framework) covers query-URL encoding, threshold colouring, and envelope grouping. - Verified live: docker build + container against kb-postgres@PIHA over LAN and Ollama@SOLARIA over Tailscale -- GET / (HTML), /static/app.js, /healthz, and /search (cascade + flat) all round-tripped correctly, including real summary/summary_tags data. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:39:46 +02:00
BASE_DIR = pathlib.Path(__file__).resolve().parent
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
KB_DSN = os.environ.get("KB_DSN")
feat(kb): aktywny fallback embeddingów SOLARIA→PIHA dla kb-query (faza 4 Krok 2) Ostatni krok fazy 4 KB (plan §2 Decyzja 2, §5): kb-query przestaje być martwe przez ~16 h/dobę, gdy SOLARIA (GPU) śpi — zapytania embeduje wtedy lokalna Ollama CPU na PIHA (wolniej: ~790 ms+ vs ~207 ms na GPU, ale działa). Nowy serwis services/ollama-piha (GitOps, owner_node: piha): - ollama/ollama:latest (arm64 natywnie), OLLAMA_KEEP_ALIVE=0 — model zwalnia RAM natychmiast po każdym wywołaniu (spike, nie rezydent; PIHA dzieli 8 GB z HA) - bind wyłącznie 127.0.0.1 + LAN_BIND_IP (192.168.31.5), nigdy 0.0.0.0/Tailscale - named volume ollama_piha_models (NVMe data-root) zamiast bind-mounta — obraz biega jako root w kontenerze i bind łamałby wzorzec uid PIHA (oskar=1004, kontenery uid 1000, setgid pi) - override hosts/piha/runtime/ollama-piha: mem_limit 2560m (wartość startowa z planu, do potwierdzenia kalibracją na żywo), świadomie bez mem_reservation - pull bge-m3 to jawny, ręczny krok deployu (README) — obraz nie ma modeli kb-query — maszyna stanów fallbacku (app/embed_router.py): - health-check SOLARII (GET /api/tags, timeout 1.5 s) z cache 30 s — zero sondowania per request; po powrocie SOLARII ruch wraca na GPU w ≤30 s - primary up → embed na SOLARII z twardym timeoutem 3 s; błąd W TRAKCIE zapytania = jednorazowe przełączenie (krok 3b planu): status down na 30 s i TO SAMO zapytanie leci na fallback — user nie widzi błędu SOLARII - primary down → embed prosto na ollama-piha (bez twardego timeoutu: CPU + zimny load modelu to legalnie pojedyncze sekundy) - 503 tylko gdy oba backendy padłe (lub fallback nieskonfigurowany) - inwariant modelu, druga połowa: każdy backend weryfikowany raz, leniwie przy pierwszym użyciu, że /api/tags zawiera EMBED_MODEL (bge-m3 — ta sama wartość co startowy check przeciw document_chunk.model/document_summary.embedding_model); niezgodność = ERROR log + 500, nigdy ciche liczenie dystansów między różnymi przestrzeniami embeddingów; leniwie, bo śpiąca SOLARIA nie może blokować startu serwisu - odpowiedź /search: nowe pole embed_backend ("solaria"|"piha") + sol_status wg realnego świata routera (UI już renderuje down jako "offline (fallback embed)"); log INFO backend=... elapsed_ms=... per zapytanie - /healthz: sol_status przez cache routera (spójny widok z routingiem) + fallback_status (żywa, tania sonda /api/tags) Konfiguracja spójnie przez env (compose + env.example + service.yaml + README): EMBED_PRIMARY_URL (zastępuje OLLAMA_URL), EMBED_FALLBACK_URL (pusty = brak fallbacku, zachowanie sprzed kroku 2), EMBED_{PRIMARY,FALLBACK}_NAME, EMBED_HEALTH_TTL_S/EMBED_HEALTH_TIMEOUT_S/EMBED_PRIMARY_TIMEOUT_S. Testy: 39 pass (14 nowych w test_embed_router.py: cache TTL, failover w trakcie zapytania, powrót po TTL, oba padłe, mismatch modelu na primary i fallbacku, tag "bge-m3:latest" vs "bge-m3"); docker build + smoke (importy + uvicorn do guardu KB_DSN) OK; compose config OK dla obu stacków. Deploy (Oskar, na PIHA z mastera po merge): cd ~/homelab-codex-ws && git pull # 1. ollama-piha cp services/ollama-piha/env.example services/ollama-piha/.env docker compose -f services/ollama-piha/docker-compose.yml \ -f hosts/piha/runtime/ollama-piha/docker-compose.override.yml \ --env-file services/ollama-piha/.env up -d docker exec ollama-piha ollama pull bge-m3 # ręczny krok, obowiązkowy services/ollama-piha/healthcheck.sh # 2. kb-query (dopisać fallback do istniejącego .env) echo 'EMBED_FALLBACK_URL=http://192.168.31.5:11434' >> services/kb-query/.env docker compose -f services/kb-query/docker-compose.yml \ -f hosts/piha/runtime/kb-query/docker-compose.override.yml up -d --build services/kb-query/healthcheck.sh # (deploy-node.sh też podniesie oba serwisy z hosts/piha/services.yaml, # ale pull bge-m3 i .env pozostają ręczne) Weryfikacja: testy A/B/C w services/kb-query/README.md (backend=solaria przy SOLARII online; backend=piha przy symulacji offline; powrót na GPU w ≤30 s). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-29 19:01:29 +02:00
# EMBED_PRIMARY_URL replaced OLLAMA_URL in Krok 2 (aktywny fallback) -- one naming scheme
# for both legs. Same default upstream as before: Ollama @ SOLARIA over Tailscale.
EMBED_PRIMARY_URL = os.environ.get("EMBED_PRIMARY_URL", "http://solaria:11434")
# Unset/empty -> no fallback leg: /search 503s when SOLARIA is down (pre-Krok-2 behaviour).
EMBED_FALLBACK_URL = os.environ.get("EMBED_FALLBACK_URL") or None
EMBED_PRIMARY_NAME = os.environ.get("EMBED_PRIMARY_NAME", "solaria")
EMBED_FALLBACK_NAME = os.environ.get("EMBED_FALLBACK_NAME", "piha")
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
EMBED_MODEL = os.environ.get("EMBED_MODEL", "bge-m3")
SUMMARY_MODEL = os.environ.get("SUMMARY_MODEL", "claude-haiku-4-5")
feat(kb): aktywny fallback embeddingów SOLARIA→PIHA dla kb-query (faza 4 Krok 2) Ostatni krok fazy 4 KB (plan §2 Decyzja 2, §5): kb-query przestaje być martwe przez ~16 h/dobę, gdy SOLARIA (GPU) śpi — zapytania embeduje wtedy lokalna Ollama CPU na PIHA (wolniej: ~790 ms+ vs ~207 ms na GPU, ale działa). Nowy serwis services/ollama-piha (GitOps, owner_node: piha): - ollama/ollama:latest (arm64 natywnie), OLLAMA_KEEP_ALIVE=0 — model zwalnia RAM natychmiast po każdym wywołaniu (spike, nie rezydent; PIHA dzieli 8 GB z HA) - bind wyłącznie 127.0.0.1 + LAN_BIND_IP (192.168.31.5), nigdy 0.0.0.0/Tailscale - named volume ollama_piha_models (NVMe data-root) zamiast bind-mounta — obraz biega jako root w kontenerze i bind łamałby wzorzec uid PIHA (oskar=1004, kontenery uid 1000, setgid pi) - override hosts/piha/runtime/ollama-piha: mem_limit 2560m (wartość startowa z planu, do potwierdzenia kalibracją na żywo), świadomie bez mem_reservation - pull bge-m3 to jawny, ręczny krok deployu (README) — obraz nie ma modeli kb-query — maszyna stanów fallbacku (app/embed_router.py): - health-check SOLARII (GET /api/tags, timeout 1.5 s) z cache 30 s — zero sondowania per request; po powrocie SOLARII ruch wraca na GPU w ≤30 s - primary up → embed na SOLARII z twardym timeoutem 3 s; błąd W TRAKCIE zapytania = jednorazowe przełączenie (krok 3b planu): status down na 30 s i TO SAMO zapytanie leci na fallback — user nie widzi błędu SOLARII - primary down → embed prosto na ollama-piha (bez twardego timeoutu: CPU + zimny load modelu to legalnie pojedyncze sekundy) - 503 tylko gdy oba backendy padłe (lub fallback nieskonfigurowany) - inwariant modelu, druga połowa: każdy backend weryfikowany raz, leniwie przy pierwszym użyciu, że /api/tags zawiera EMBED_MODEL (bge-m3 — ta sama wartość co startowy check przeciw document_chunk.model/document_summary.embedding_model); niezgodność = ERROR log + 500, nigdy ciche liczenie dystansów między różnymi przestrzeniami embeddingów; leniwie, bo śpiąca SOLARIA nie może blokować startu serwisu - odpowiedź /search: nowe pole embed_backend ("solaria"|"piha") + sol_status wg realnego świata routera (UI już renderuje down jako "offline (fallback embed)"); log INFO backend=... elapsed_ms=... per zapytanie - /healthz: sol_status przez cache routera (spójny widok z routingiem) + fallback_status (żywa, tania sonda /api/tags) Konfiguracja spójnie przez env (compose + env.example + service.yaml + README): EMBED_PRIMARY_URL (zastępuje OLLAMA_URL), EMBED_FALLBACK_URL (pusty = brak fallbacku, zachowanie sprzed kroku 2), EMBED_{PRIMARY,FALLBACK}_NAME, EMBED_HEALTH_TTL_S/EMBED_HEALTH_TIMEOUT_S/EMBED_PRIMARY_TIMEOUT_S. Testy: 39 pass (14 nowych w test_embed_router.py: cache TTL, failover w trakcie zapytania, powrót po TTL, oba padłe, mismatch modelu na primary i fallbacku, tag "bge-m3:latest" vs "bge-m3"); docker build + smoke (importy + uvicorn do guardu KB_DSN) OK; compose config OK dla obu stacków. Deploy (Oskar, na PIHA z mastera po merge): cd ~/homelab-codex-ws && git pull # 1. ollama-piha cp services/ollama-piha/env.example services/ollama-piha/.env docker compose -f services/ollama-piha/docker-compose.yml \ -f hosts/piha/runtime/ollama-piha/docker-compose.override.yml \ --env-file services/ollama-piha/.env up -d docker exec ollama-piha ollama pull bge-m3 # ręczny krok, obowiązkowy services/ollama-piha/healthcheck.sh # 2. kb-query (dopisać fallback do istniejącego .env) echo 'EMBED_FALLBACK_URL=http://192.168.31.5:11434' >> services/kb-query/.env docker compose -f services/kb-query/docker-compose.yml \ -f hosts/piha/runtime/kb-query/docker-compose.override.yml up -d --build services/kb-query/healthcheck.sh # (deploy-node.sh też podniesie oba serwisy z hosts/piha/services.yaml, # ale pull bge-m3 i .env pozostają ręczne) Weryfikacja: testy A/B/C w services/kb-query/README.md (backend=solaria przy SOLARII online; backend=piha przy symulacji offline; powrót na GPU w ≤30 s). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-29 19:01:29 +02:00
# State-machine parameters (plan §2 D2 table; timeouts rationale in app/embed_router.py).
EMBED_HEALTH_TTL_S = float(os.environ.get("EMBED_HEALTH_TTL_S", "30"))
EMBED_HEALTH_TIMEOUT_S = float(os.environ.get("EMBED_HEALTH_TIMEOUT_S", "1.5"))
EMBED_PRIMARY_TIMEOUT_S = float(os.environ.get("EMBED_PRIMARY_TIMEOUT_S", "3"))
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
@asynccontextmanager
async def lifespan(app: FastAPI):
if not KB_DSN:
raise RuntimeError("KB_DSN is required (see env.example)")
pool = await create_pool(KB_DSN)
async with pool.acquire() as conn:
# Hard invariant (plan §2 decision 2): refuse to start rather than silently serve
feat(kb): aktywny fallback embeddingów SOLARIA→PIHA dla kb-query (faza 4 Krok 2) Ostatni krok fazy 4 KB (plan §2 Decyzja 2, §5): kb-query przestaje być martwe przez ~16 h/dobę, gdy SOLARIA (GPU) śpi — zapytania embeduje wtedy lokalna Ollama CPU na PIHA (wolniej: ~790 ms+ vs ~207 ms na GPU, ale działa). Nowy serwis services/ollama-piha (GitOps, owner_node: piha): - ollama/ollama:latest (arm64 natywnie), OLLAMA_KEEP_ALIVE=0 — model zwalnia RAM natychmiast po każdym wywołaniu (spike, nie rezydent; PIHA dzieli 8 GB z HA) - bind wyłącznie 127.0.0.1 + LAN_BIND_IP (192.168.31.5), nigdy 0.0.0.0/Tailscale - named volume ollama_piha_models (NVMe data-root) zamiast bind-mounta — obraz biega jako root w kontenerze i bind łamałby wzorzec uid PIHA (oskar=1004, kontenery uid 1000, setgid pi) - override hosts/piha/runtime/ollama-piha: mem_limit 2560m (wartość startowa z planu, do potwierdzenia kalibracją na żywo), świadomie bez mem_reservation - pull bge-m3 to jawny, ręczny krok deployu (README) — obraz nie ma modeli kb-query — maszyna stanów fallbacku (app/embed_router.py): - health-check SOLARII (GET /api/tags, timeout 1.5 s) z cache 30 s — zero sondowania per request; po powrocie SOLARII ruch wraca na GPU w ≤30 s - primary up → embed na SOLARII z twardym timeoutem 3 s; błąd W TRAKCIE zapytania = jednorazowe przełączenie (krok 3b planu): status down na 30 s i TO SAMO zapytanie leci na fallback — user nie widzi błędu SOLARII - primary down → embed prosto na ollama-piha (bez twardego timeoutu: CPU + zimny load modelu to legalnie pojedyncze sekundy) - 503 tylko gdy oba backendy padłe (lub fallback nieskonfigurowany) - inwariant modelu, druga połowa: każdy backend weryfikowany raz, leniwie przy pierwszym użyciu, że /api/tags zawiera EMBED_MODEL (bge-m3 — ta sama wartość co startowy check przeciw document_chunk.model/document_summary.embedding_model); niezgodność = ERROR log + 500, nigdy ciche liczenie dystansów między różnymi przestrzeniami embeddingów; leniwie, bo śpiąca SOLARIA nie może blokować startu serwisu - odpowiedź /search: nowe pole embed_backend ("solaria"|"piha") + sol_status wg realnego świata routera (UI już renderuje down jako "offline (fallback embed)"); log INFO backend=... elapsed_ms=... per zapytanie - /healthz: sol_status przez cache routera (spójny widok z routingiem) + fallback_status (żywa, tania sonda /api/tags) Konfiguracja spójnie przez env (compose + env.example + service.yaml + README): EMBED_PRIMARY_URL (zastępuje OLLAMA_URL), EMBED_FALLBACK_URL (pusty = brak fallbacku, zachowanie sprzed kroku 2), EMBED_{PRIMARY,FALLBACK}_NAME, EMBED_HEALTH_TTL_S/EMBED_HEALTH_TIMEOUT_S/EMBED_PRIMARY_TIMEOUT_S. Testy: 39 pass (14 nowych w test_embed_router.py: cache TTL, failover w trakcie zapytania, powrót po TTL, oba padłe, mismatch modelu na primary i fallbacku, tag "bge-m3:latest" vs "bge-m3"); docker build + smoke (importy + uvicorn do guardu KB_DSN) OK; compose config OK dla obu stacków. Deploy (Oskar, na PIHA z mastera po merge): cd ~/homelab-codex-ws && git pull # 1. ollama-piha cp services/ollama-piha/env.example services/ollama-piha/.env docker compose -f services/ollama-piha/docker-compose.yml \ -f hosts/piha/runtime/ollama-piha/docker-compose.override.yml \ --env-file services/ollama-piha/.env up -d docker exec ollama-piha ollama pull bge-m3 # ręczny krok, obowiązkowy services/ollama-piha/healthcheck.sh # 2. kb-query (dopisać fallback do istniejącego .env) echo 'EMBED_FALLBACK_URL=http://192.168.31.5:11434' >> services/kb-query/.env docker compose -f services/kb-query/docker-compose.yml \ -f hosts/piha/runtime/kb-query/docker-compose.override.yml up -d --build services/kb-query/healthcheck.sh # (deploy-node.sh też podniesie oba serwisy z hosts/piha/services.yaml, # ale pull bge-m3 i .env pozostają ręczne) Weryfikacja: testy A/B/C w services/kb-query/README.md (backend=solaria przy SOLARII online; backend=piha przy symulacji offline; powrót na GPU w ≤30 s). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-29 19:01:29 +02:00
# queries against a mismatched embedding space. The per-backend half of the same
# invariant (does the Ollama actually serve EMBED_MODEL?) lives in EmbedRouter and
# runs lazily at each backend's first use -- a sleeping SOLARIA must not block boot.
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
await validate_embed_model(conn, EMBED_MODEL)
app.state.pool = pool
app.state.http = aiohttp.ClientSession()
feat(kb): aktywny fallback embeddingów SOLARIA→PIHA dla kb-query (faza 4 Krok 2) Ostatni krok fazy 4 KB (plan §2 Decyzja 2, §5): kb-query przestaje być martwe przez ~16 h/dobę, gdy SOLARIA (GPU) śpi — zapytania embeduje wtedy lokalna Ollama CPU na PIHA (wolniej: ~790 ms+ vs ~207 ms na GPU, ale działa). Nowy serwis services/ollama-piha (GitOps, owner_node: piha): - ollama/ollama:latest (arm64 natywnie), OLLAMA_KEEP_ALIVE=0 — model zwalnia RAM natychmiast po każdym wywołaniu (spike, nie rezydent; PIHA dzieli 8 GB z HA) - bind wyłącznie 127.0.0.1 + LAN_BIND_IP (192.168.31.5), nigdy 0.0.0.0/Tailscale - named volume ollama_piha_models (NVMe data-root) zamiast bind-mounta — obraz biega jako root w kontenerze i bind łamałby wzorzec uid PIHA (oskar=1004, kontenery uid 1000, setgid pi) - override hosts/piha/runtime/ollama-piha: mem_limit 2560m (wartość startowa z planu, do potwierdzenia kalibracją na żywo), świadomie bez mem_reservation - pull bge-m3 to jawny, ręczny krok deployu (README) — obraz nie ma modeli kb-query — maszyna stanów fallbacku (app/embed_router.py): - health-check SOLARII (GET /api/tags, timeout 1.5 s) z cache 30 s — zero sondowania per request; po powrocie SOLARII ruch wraca na GPU w ≤30 s - primary up → embed na SOLARII z twardym timeoutem 3 s; błąd W TRAKCIE zapytania = jednorazowe przełączenie (krok 3b planu): status down na 30 s i TO SAMO zapytanie leci na fallback — user nie widzi błędu SOLARII - primary down → embed prosto na ollama-piha (bez twardego timeoutu: CPU + zimny load modelu to legalnie pojedyncze sekundy) - 503 tylko gdy oba backendy padłe (lub fallback nieskonfigurowany) - inwariant modelu, druga połowa: każdy backend weryfikowany raz, leniwie przy pierwszym użyciu, że /api/tags zawiera EMBED_MODEL (bge-m3 — ta sama wartość co startowy check przeciw document_chunk.model/document_summary.embedding_model); niezgodność = ERROR log + 500, nigdy ciche liczenie dystansów między różnymi przestrzeniami embeddingów; leniwie, bo śpiąca SOLARIA nie może blokować startu serwisu - odpowiedź /search: nowe pole embed_backend ("solaria"|"piha") + sol_status wg realnego świata routera (UI już renderuje down jako "offline (fallback embed)"); log INFO backend=... elapsed_ms=... per zapytanie - /healthz: sol_status przez cache routera (spójny widok z routingiem) + fallback_status (żywa, tania sonda /api/tags) Konfiguracja spójnie przez env (compose + env.example + service.yaml + README): EMBED_PRIMARY_URL (zastępuje OLLAMA_URL), EMBED_FALLBACK_URL (pusty = brak fallbacku, zachowanie sprzed kroku 2), EMBED_{PRIMARY,FALLBACK}_NAME, EMBED_HEALTH_TTL_S/EMBED_HEALTH_TIMEOUT_S/EMBED_PRIMARY_TIMEOUT_S. Testy: 39 pass (14 nowych w test_embed_router.py: cache TTL, failover w trakcie zapytania, powrót po TTL, oba padłe, mismatch modelu na primary i fallbacku, tag "bge-m3:latest" vs "bge-m3"); docker build + smoke (importy + uvicorn do guardu KB_DSN) OK; compose config OK dla obu stacków. Deploy (Oskar, na PIHA z mastera po merge): cd ~/homelab-codex-ws && git pull # 1. ollama-piha cp services/ollama-piha/env.example services/ollama-piha/.env docker compose -f services/ollama-piha/docker-compose.yml \ -f hosts/piha/runtime/ollama-piha/docker-compose.override.yml \ --env-file services/ollama-piha/.env up -d docker exec ollama-piha ollama pull bge-m3 # ręczny krok, obowiązkowy services/ollama-piha/healthcheck.sh # 2. kb-query (dopisać fallback do istniejącego .env) echo 'EMBED_FALLBACK_URL=http://192.168.31.5:11434' >> services/kb-query/.env docker compose -f services/kb-query/docker-compose.yml \ -f hosts/piha/runtime/kb-query/docker-compose.override.yml up -d --build services/kb-query/healthcheck.sh # (deploy-node.sh też podniesie oba serwisy z hosts/piha/services.yaml, # ale pull bge-m3 i .env pozostają ręczne) Weryfikacja: testy A/B/C w services/kb-query/README.md (backend=solaria przy SOLARII online; backend=piha przy symulacji offline; powrót na GPU w ≤30 s). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-29 19:01:29 +02:00
app.state.embed_router = EmbedRouter(
EMBED_PRIMARY_URL,
EMBED_FALLBACK_URL,
embed_model=EMBED_MODEL,
primary_name=EMBED_PRIMARY_NAME,
fallback_name=EMBED_FALLBACK_NAME,
health_ttl_s=EMBED_HEALTH_TTL_S,
health_timeout_s=EMBED_HEALTH_TIMEOUT_S,
primary_embed_timeout_s=EMBED_PRIMARY_TIMEOUT_S,
)
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
try:
yield
finally:
await app.state.http.close()
await pool.close()
app = FastAPI(lifespan=lifespan)
feat(kb-query): add search frontend (module 5 phase 4, plan §7, Krok 4) Krok 4 of the phase-4 plan done ahead of the local-embed-fallback step (Krok 2, deliberately deferred -- embed stays a plain SOLARIA call, per task instruction): one FastAPI process now serves both the /search API and the UI, no separate frontend build (plan §2 decision 4). - GET / renders a Jinja2 shell; app/static/app.js (vanilla, no build) and style.css are the whole client. Query -> /search, results grouped by envelope_id client-side (chunks sorted by dist, <details> fragments). - Colour thresholds per plan §7: dist<0.45 green, 0.45-0.55 yellow (still shown with a warning), >0.55 never rendered as an individual result; if a query ends up with nothing renderable, one "Brak odpowiedzi w KB" message replaces the list, carrying the best observed dist. - Paperless hits link out; gmail hits get a "kopiuj Message-ID" button (there's nothing to link to yet, plan §2 decision 3) plus header metadata. Cascade/flat toggle defaults to cascade. Footer shows sol_status, refreshed from /healthz on load and after each search. - /search gained additive summary/summary_tags fields (document_summary, haiku track) so the UI can show a document summary as each result group's header -- non-breaking, existing response shape untouched. - Tests: app/db.py + app/search.py unit tests (mocked DB/HTTP, no live deps) cover the new summary join; tests/test_frontend.py drives GET / and /static/* via TestClient without running the DB-requiring lifespan; tests/frontend/app.test.js (Node's built-in test runner, no framework) covers query-URL encoding, threshold colouring, and envelope grouping. - Verified live: docker build + container against kb-postgres@PIHA over LAN and Ollama@SOLARIA over Tailscale -- GET / (HTML), /static/app.js, /healthz, and /search (cascade + flat) all round-tripped correctly, including real summary/summary_tags data. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:39:46 +02:00
app.mount("/static", StaticFiles(directory=BASE_DIR / "static"), name="static")
templates = Jinja2Templates(directory=BASE_DIR / "templates")
@app.get("/", response_class=HTMLResponse)
async def index(request: Request):
return templates.TemplateResponse(request, "index.html")
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
@app.get("/healthz")
async def healthz() -> dict:
feat(kb): aktywny fallback embeddingów SOLARIA→PIHA dla kb-query (faza 4 Krok 2) Ostatni krok fazy 4 KB (plan §2 Decyzja 2, §5): kb-query przestaje być martwe przez ~16 h/dobę, gdy SOLARIA (GPU) śpi — zapytania embeduje wtedy lokalna Ollama CPU na PIHA (wolniej: ~790 ms+ vs ~207 ms na GPU, ale działa). Nowy serwis services/ollama-piha (GitOps, owner_node: piha): - ollama/ollama:latest (arm64 natywnie), OLLAMA_KEEP_ALIVE=0 — model zwalnia RAM natychmiast po każdym wywołaniu (spike, nie rezydent; PIHA dzieli 8 GB z HA) - bind wyłącznie 127.0.0.1 + LAN_BIND_IP (192.168.31.5), nigdy 0.0.0.0/Tailscale - named volume ollama_piha_models (NVMe data-root) zamiast bind-mounta — obraz biega jako root w kontenerze i bind łamałby wzorzec uid PIHA (oskar=1004, kontenery uid 1000, setgid pi) - override hosts/piha/runtime/ollama-piha: mem_limit 2560m (wartość startowa z planu, do potwierdzenia kalibracją na żywo), świadomie bez mem_reservation - pull bge-m3 to jawny, ręczny krok deployu (README) — obraz nie ma modeli kb-query — maszyna stanów fallbacku (app/embed_router.py): - health-check SOLARII (GET /api/tags, timeout 1.5 s) z cache 30 s — zero sondowania per request; po powrocie SOLARII ruch wraca na GPU w ≤30 s - primary up → embed na SOLARII z twardym timeoutem 3 s; błąd W TRAKCIE zapytania = jednorazowe przełączenie (krok 3b planu): status down na 30 s i TO SAMO zapytanie leci na fallback — user nie widzi błędu SOLARII - primary down → embed prosto na ollama-piha (bez twardego timeoutu: CPU + zimny load modelu to legalnie pojedyncze sekundy) - 503 tylko gdy oba backendy padłe (lub fallback nieskonfigurowany) - inwariant modelu, druga połowa: każdy backend weryfikowany raz, leniwie przy pierwszym użyciu, że /api/tags zawiera EMBED_MODEL (bge-m3 — ta sama wartość co startowy check przeciw document_chunk.model/document_summary.embedding_model); niezgodność = ERROR log + 500, nigdy ciche liczenie dystansów między różnymi przestrzeniami embeddingów; leniwie, bo śpiąca SOLARIA nie może blokować startu serwisu - odpowiedź /search: nowe pole embed_backend ("solaria"|"piha") + sol_status wg realnego świata routera (UI już renderuje down jako "offline (fallback embed)"); log INFO backend=... elapsed_ms=... per zapytanie - /healthz: sol_status przez cache routera (spójny widok z routingiem) + fallback_status (żywa, tania sonda /api/tags) Konfiguracja spójnie przez env (compose + env.example + service.yaml + README): EMBED_PRIMARY_URL (zastępuje OLLAMA_URL), EMBED_FALLBACK_URL (pusty = brak fallbacku, zachowanie sprzed kroku 2), EMBED_{PRIMARY,FALLBACK}_NAME, EMBED_HEALTH_TTL_S/EMBED_HEALTH_TIMEOUT_S/EMBED_PRIMARY_TIMEOUT_S. Testy: 39 pass (14 nowych w test_embed_router.py: cache TTL, failover w trakcie zapytania, powrót po TTL, oba padłe, mismatch modelu na primary i fallbacku, tag "bge-m3:latest" vs "bge-m3"); docker build + smoke (importy + uvicorn do guardu KB_DSN) OK; compose config OK dla obu stacków. Deploy (Oskar, na PIHA z mastera po merge): cd ~/homelab-codex-ws && git pull # 1. ollama-piha cp services/ollama-piha/env.example services/ollama-piha/.env docker compose -f services/ollama-piha/docker-compose.yml \ -f hosts/piha/runtime/ollama-piha/docker-compose.override.yml \ --env-file services/ollama-piha/.env up -d docker exec ollama-piha ollama pull bge-m3 # ręczny krok, obowiązkowy services/ollama-piha/healthcheck.sh # 2. kb-query (dopisać fallback do istniejącego .env) echo 'EMBED_FALLBACK_URL=http://192.168.31.5:11434' >> services/kb-query/.env docker compose -f services/kb-query/docker-compose.yml \ -f hosts/piha/runtime/kb-query/docker-compose.override.yml up -d --build services/kb-query/healthcheck.sh # (deploy-node.sh też podniesie oba serwisy z hosts/piha/services.yaml, # ale pull bge-m3 i .env pozostają ręczne) Weryfikacja: testy A/B/C w services/kb-query/README.md (backend=solaria przy SOLARII online; backend=piha przy symulacji offline; powrót na GPU w ≤30 s). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-29 19:01:29 +02:00
# sol_status goes through the router's 30 s cache -- /healthz and /search routing
# deliberately share one world view (and monitoring no longer re-probes a sleeping
# SOLARIA on every scrape). fallback_status is a live cheap /api/tags probe.
router = app.state.embed_router
return {
"status": "ok",
"sol_status": await router.primary_status(app.state.http),
"fallback_status": await router.fallback_status(app.state.http),
}
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
@app.get("/search")
async def search(
q: str = Query(..., min_length=1),
feat(kb-query): domyslny mode /search = hybrid (DoD (d) fazy mailowej) Eval na pelnym korpusie 2026-08-06 (187 025 zembedowanych chunkow mailowych w HNSW) dal PASS: kryterium 1 (regresja paperless) bez degradacji zadnego istniejacego hitu we flat ani w hybrid, mailowe hit@3 5/5. Koszt hybrydy to jedno dodatkowe zapytanie SQL na wyszukiwanie. Surowe wyniki: eval-http-2026-08-06.json / eval-direct-2026-08-06.json w ~/kb/mail/ingest-logs na PIHA (niecommitowane, artefakt runu). - app/main.py: Query("cascade") -> Query("hybrid"); pattern bez zmian, wiec jawne ?mode=cascade i ?mode=flat dzialaja dokladnie jak dotad. - app/static/app.js: przy odznaczonym "tryb flat (debug)" UI nie wysyla juz parametru mode w ogole -- dziedziczy default API. Default zdefiniowany w jednym miejscu (serwer), nie zduplikowany w JS. - testy: nowa klasa TestSearchEndpointModeDefault (TestClient bez lifespan, fake pool/router) sprawdza kontrakt HTTP -- brak mode => tor hybrid (weryfikowany po obecnosci koperty gmail osiagalnej wylacznie galezia hybrid, nie po samej etykiecie), jawne mode=flat / mode=cascade => stare tory, nieznany mode => 422. Frontend: buildSearchUrl pomija mode gdy brak. - docs: kb/services/kb-query.md (tabela trybow + endpoint + przyklad odpowiedzi + opis przelacznika w UI), env.example/service.yaml (komentarze SUMMARY_MODEL; default mode nie jest konfigurowalny przez env), kb/phases/kb-m5-faza-mailowa.md (DoD (d) SPELNIONE 2026-08-06 + wzmianki w Kroku 3, Wyniku bramki, decyzjach Etapu B i tabeli planu). Weryfikacja: pytest services/kb-query -> 46 passed; node --test tests/frontend/app.test.js -> 6/6; docker build OK + smoke run (uvicorn startuje, bez KB_DSN swiadomie konczy sie RuntimeError z env.example). Deploy NIE wykonany -- operator wdraza z mastera na PIHA po mergu. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-06 13:30:02 +02:00
mode: str = Query("hybrid", pattern="^(cascade|flat|hybrid)$"),
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
) -> dict:
try:
async with app.state.pool.acquire() as conn:
return await run_search(
feat(kb): aktywny fallback embeddingów SOLARIA→PIHA dla kb-query (faza 4 Krok 2) Ostatni krok fazy 4 KB (plan §2 Decyzja 2, §5): kb-query przestaje być martwe przez ~16 h/dobę, gdy SOLARIA (GPU) śpi — zapytania embeduje wtedy lokalna Ollama CPU na PIHA (wolniej: ~790 ms+ vs ~207 ms na GPU, ale działa). Nowy serwis services/ollama-piha (GitOps, owner_node: piha): - ollama/ollama:latest (arm64 natywnie), OLLAMA_KEEP_ALIVE=0 — model zwalnia RAM natychmiast po każdym wywołaniu (spike, nie rezydent; PIHA dzieli 8 GB z HA) - bind wyłącznie 127.0.0.1 + LAN_BIND_IP (192.168.31.5), nigdy 0.0.0.0/Tailscale - named volume ollama_piha_models (NVMe data-root) zamiast bind-mounta — obraz biega jako root w kontenerze i bind łamałby wzorzec uid PIHA (oskar=1004, kontenery uid 1000, setgid pi) - override hosts/piha/runtime/ollama-piha: mem_limit 2560m (wartość startowa z planu, do potwierdzenia kalibracją na żywo), świadomie bez mem_reservation - pull bge-m3 to jawny, ręczny krok deployu (README) — obraz nie ma modeli kb-query — maszyna stanów fallbacku (app/embed_router.py): - health-check SOLARII (GET /api/tags, timeout 1.5 s) z cache 30 s — zero sondowania per request; po powrocie SOLARII ruch wraca na GPU w ≤30 s - primary up → embed na SOLARII z twardym timeoutem 3 s; błąd W TRAKCIE zapytania = jednorazowe przełączenie (krok 3b planu): status down na 30 s i TO SAMO zapytanie leci na fallback — user nie widzi błędu SOLARII - primary down → embed prosto na ollama-piha (bez twardego timeoutu: CPU + zimny load modelu to legalnie pojedyncze sekundy) - 503 tylko gdy oba backendy padłe (lub fallback nieskonfigurowany) - inwariant modelu, druga połowa: każdy backend weryfikowany raz, leniwie przy pierwszym użyciu, że /api/tags zawiera EMBED_MODEL (bge-m3 — ta sama wartość co startowy check przeciw document_chunk.model/document_summary.embedding_model); niezgodność = ERROR log + 500, nigdy ciche liczenie dystansów między różnymi przestrzeniami embeddingów; leniwie, bo śpiąca SOLARIA nie może blokować startu serwisu - odpowiedź /search: nowe pole embed_backend ("solaria"|"piha") + sol_status wg realnego świata routera (UI już renderuje down jako "offline (fallback embed)"); log INFO backend=... elapsed_ms=... per zapytanie - /healthz: sol_status przez cache routera (spójny widok z routingiem) + fallback_status (żywa, tania sonda /api/tags) Konfiguracja spójnie przez env (compose + env.example + service.yaml + README): EMBED_PRIMARY_URL (zastępuje OLLAMA_URL), EMBED_FALLBACK_URL (pusty = brak fallbacku, zachowanie sprzed kroku 2), EMBED_{PRIMARY,FALLBACK}_NAME, EMBED_HEALTH_TTL_S/EMBED_HEALTH_TIMEOUT_S/EMBED_PRIMARY_TIMEOUT_S. Testy: 39 pass (14 nowych w test_embed_router.py: cache TTL, failover w trakcie zapytania, powrót po TTL, oba padłe, mismatch modelu na primary i fallbacku, tag "bge-m3:latest" vs "bge-m3"); docker build + smoke (importy + uvicorn do guardu KB_DSN) OK; compose config OK dla obu stacków. Deploy (Oskar, na PIHA z mastera po merge): cd ~/homelab-codex-ws && git pull # 1. ollama-piha cp services/ollama-piha/env.example services/ollama-piha/.env docker compose -f services/ollama-piha/docker-compose.yml \ -f hosts/piha/runtime/ollama-piha/docker-compose.override.yml \ --env-file services/ollama-piha/.env up -d docker exec ollama-piha ollama pull bge-m3 # ręczny krok, obowiązkowy services/ollama-piha/healthcheck.sh # 2. kb-query (dopisać fallback do istniejącego .env) echo 'EMBED_FALLBACK_URL=http://192.168.31.5:11434' >> services/kb-query/.env docker compose -f services/kb-query/docker-compose.yml \ -f hosts/piha/runtime/kb-query/docker-compose.override.yml up -d --build services/kb-query/healthcheck.sh # (deploy-node.sh też podniesie oba serwisy z hosts/piha/services.yaml, # ale pull bge-m3 i .env pozostają ręczne) Weryfikacja: testy A/B/C w services/kb-query/README.md (backend=solaria przy SOLARII online; backend=piha przy symulacji offline; powrót na GPU w ≤30 s). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-29 19:01:29 +02:00
conn, app.state.http, app.state.embed_router, q, mode, SUMMARY_MODEL
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
)
feat(kb): aktywny fallback embeddingów SOLARIA→PIHA dla kb-query (faza 4 Krok 2) Ostatni krok fazy 4 KB (plan §2 Decyzja 2, §5): kb-query przestaje być martwe przez ~16 h/dobę, gdy SOLARIA (GPU) śpi — zapytania embeduje wtedy lokalna Ollama CPU na PIHA (wolniej: ~790 ms+ vs ~207 ms na GPU, ale działa). Nowy serwis services/ollama-piha (GitOps, owner_node: piha): - ollama/ollama:latest (arm64 natywnie), OLLAMA_KEEP_ALIVE=0 — model zwalnia RAM natychmiast po każdym wywołaniu (spike, nie rezydent; PIHA dzieli 8 GB z HA) - bind wyłącznie 127.0.0.1 + LAN_BIND_IP (192.168.31.5), nigdy 0.0.0.0/Tailscale - named volume ollama_piha_models (NVMe data-root) zamiast bind-mounta — obraz biega jako root w kontenerze i bind łamałby wzorzec uid PIHA (oskar=1004, kontenery uid 1000, setgid pi) - override hosts/piha/runtime/ollama-piha: mem_limit 2560m (wartość startowa z planu, do potwierdzenia kalibracją na żywo), świadomie bez mem_reservation - pull bge-m3 to jawny, ręczny krok deployu (README) — obraz nie ma modeli kb-query — maszyna stanów fallbacku (app/embed_router.py): - health-check SOLARII (GET /api/tags, timeout 1.5 s) z cache 30 s — zero sondowania per request; po powrocie SOLARII ruch wraca na GPU w ≤30 s - primary up → embed na SOLARII z twardym timeoutem 3 s; błąd W TRAKCIE zapytania = jednorazowe przełączenie (krok 3b planu): status down na 30 s i TO SAMO zapytanie leci na fallback — user nie widzi błędu SOLARII - primary down → embed prosto na ollama-piha (bez twardego timeoutu: CPU + zimny load modelu to legalnie pojedyncze sekundy) - 503 tylko gdy oba backendy padłe (lub fallback nieskonfigurowany) - inwariant modelu, druga połowa: każdy backend weryfikowany raz, leniwie przy pierwszym użyciu, że /api/tags zawiera EMBED_MODEL (bge-m3 — ta sama wartość co startowy check przeciw document_chunk.model/document_summary.embedding_model); niezgodność = ERROR log + 500, nigdy ciche liczenie dystansów między różnymi przestrzeniami embeddingów; leniwie, bo śpiąca SOLARIA nie może blokować startu serwisu - odpowiedź /search: nowe pole embed_backend ("solaria"|"piha") + sol_status wg realnego świata routera (UI już renderuje down jako "offline (fallback embed)"); log INFO backend=... elapsed_ms=... per zapytanie - /healthz: sol_status przez cache routera (spójny widok z routingiem) + fallback_status (żywa, tania sonda /api/tags) Konfiguracja spójnie przez env (compose + env.example + service.yaml + README): EMBED_PRIMARY_URL (zastępuje OLLAMA_URL), EMBED_FALLBACK_URL (pusty = brak fallbacku, zachowanie sprzed kroku 2), EMBED_{PRIMARY,FALLBACK}_NAME, EMBED_HEALTH_TTL_S/EMBED_HEALTH_TIMEOUT_S/EMBED_PRIMARY_TIMEOUT_S. Testy: 39 pass (14 nowych w test_embed_router.py: cache TTL, failover w trakcie zapytania, powrót po TTL, oba padłe, mismatch modelu na primary i fallbacku, tag "bge-m3:latest" vs "bge-m3"); docker build + smoke (importy + uvicorn do guardu KB_DSN) OK; compose config OK dla obu stacków. Deploy (Oskar, na PIHA z mastera po merge): cd ~/homelab-codex-ws && git pull # 1. ollama-piha cp services/ollama-piha/env.example services/ollama-piha/.env docker compose -f services/ollama-piha/docker-compose.yml \ -f hosts/piha/runtime/ollama-piha/docker-compose.override.yml \ --env-file services/ollama-piha/.env up -d docker exec ollama-piha ollama pull bge-m3 # ręczny krok, obowiązkowy services/ollama-piha/healthcheck.sh # 2. kb-query (dopisać fallback do istniejącego .env) echo 'EMBED_FALLBACK_URL=http://192.168.31.5:11434' >> services/kb-query/.env docker compose -f services/kb-query/docker-compose.yml \ -f hosts/piha/runtime/kb-query/docker-compose.override.yml up -d --build services/kb-query/healthcheck.sh # (deploy-node.sh też podniesie oba serwisy z hosts/piha/services.yaml, # ale pull bge-m3 i .env pozostają ręczne) Weryfikacja: testy A/B/C w services/kb-query/README.md (backend=solaria przy SOLARII online; backend=piha przy symulacji offline; powrót na GPU w ≤30 s). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-29 19:01:29 +02:00
except ModelMismatchError as exc:
# Config/ops error (backend up but wrong model set) -- 500, deliberately loud.
raise HTTPException(status_code=500, detail=str(exc)) from exc
except (EmbedBackendError, aiohttp.ClientError) as exc:
feat(kb): add kb-query service skeleton (search API, no ingress yet) Module 5 phase 4 step 1 (docs/kb/modules/05-faza4-plan.md, §4): first user-facing HTTP entry point to the KB. FastAPI wrapping kb_retrieval.cascade_query/flat_query — GET /search (query_text -> embed via Ollama@SOLARIA -> cascade/flat -> envelope join -> JSON with per-source links) and GET /healthz. Search API only, no answer synthesis (phase 5) and no server-side dist filtering — the 0.45/0.55 colour thresholds are a frontend concern (plan §7, a later step). Hard startup invariant (plan §2 decision 2): refuses to start unless the configured EMBED_MODEL is present in both document_chunk.model and document_summary.embedding_model. Note the latter: document_summary.model is the LLM that *wrote* the summary (claude-haiku-4-5/gemma3:12b), not the embedder — checked live against kb-postgres@PIHA before writing this, see app/startup.py's docstring. Verified end-to-end with a live docker run: the invariant crash-loops on a mismatched EMBED_MODEL and passes through to a real /search hit against the live corpus with a correct model. Repo-only: no deploy, no npm/OIDC/DNS wiring (plan §8, later step), no local embed fallback (plan §5, later step) — Ollama@SOLARIA is called directly and a failure surfaces as 503, not a crash. Also: scripts/deploy/deploy.sh's gate now builds each service via `docker compose build` instead of a raw `docker build <svc_dir>`, so a service whose docker-compose.yml declares a repo-root build context (needed here to COPY packages/kb-retrieval/, the packages/ Dockerfile convention already documented in CLAUDE.md) resolves the same way in the gate as it does at real deploy time (deploy-node.sh's `docker compose ... up --build`). No behavior change for existing single-context services — verified against llm-gateway's compose file. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-22 16:06:18 +02:00
raise HTTPException(status_code=503, detail=f"embed backend unavailable: {exc}") from exc