Insert a configurable in-cluster OCR rung into the escalation ladder (Deck #353):
a tier2-eligible doc is OCR'd on the on-demand burst GPU before falling through to
paid upstream OCR. The in-cluster backend is reached ONLY via the embedding gateway
(model prefix routes to the GPU over the tailnet) and is a config value (default
surya/surya-ocr-2, swappable to e.g. lightonocr) — never hard-coded.
Ladder: fast -> structured -> ocr-incluster -> ocr-upstream
(queues ingest-ocr-incluster / ingest-ocr-upstream).
- escalation.py: 4-tier ladder; in-cluster flag folded into the dead-letter signature.
- ocr.py: OcrProcessor(name, tier, model_setting, gateway_only); build_ocr_backend(
..., model=, gateway_only=) — gateway_only forces the gateway backend (never the
direct Mistral fallback), disabling the tier with a warning if no gateway URL.
- registry.py: per-rung enable map; scanned docs target minimum="ocr-incluster";
inline path runs the cheapest available OCR rung.
- procrastinate.py: two OCR queues; legacy ingest-ocr kept as a drain target.
- config.py: DOCUMENT_OCR_INCLUSTER_ENABLED (off) + DOCUMENT_OCR_INCLUSTER_MODEL.
- __init__.py: register the two OCR instances; vector/processor.py: pages_ocr
metered for the upstream (paid) rung only; cli.py: new --tier choices + legacy drain.
- metrics.py: zero the legacy ingest-ocr queue gauge during rollout.
- tests: migrated to the split ladder + new tests (gateway-only forcing, per-tier
model incl. lightonocr override, no-hard-coded-surya guard). 1792 pass; ruff + ty green.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Round-5 review nit on PR #920 (non-blocking): record_document_dead_lettered
increments alongside the fail-safe mark_dead_letter, so the counter measures the
dead-letter attempt and can sit marginally above the live marker count if a
Qdrant write fails. Note it in the docstring.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Round-4 review nits on PR #920 (none blocking):
- record_document_dead_lettered: enumerate the oversize reason (added this PR)
alongside timeout/oom/error in the docstring + counter comment.
- Note the clear-dead-letter-before-upsert ordering implication (a transient
upsert failure re-parses once, never a silent drop).
- Clarify the orphan sweep's kept counter for tenant-wide dead-letter markers.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A pathological PDF (a 206-page ChronoScan scan with ~3400 JBIG2/JPX images)
jammed a tenant's structured ingest worker in an infinite reprocess loop,
re-burning a 120s pymupdf4llm parse (and occasionally OOM-racing the 2Gi pod)
every few minutes.
Root cause: the per-user placeholder "failed" mark could not stop the loop. The
placeholder point ID is user-agnostic (uuid5("file:<doc_id>:placeholder")) but
the scanner's freshness gate, query, and status update all filter by user_id.
For a file visible to several users the single shared placeholder's user_id is
overwritten by whoever scanned last, so every other user's scan sees "no record"
and re-queues -- an N-user ping-pong that never honours the failed status.
Fix: when a parse fails terminally (no higher escalation tier available, e.g.
structured with OCR off) record a durable, content-addressed, user-agnostic
dead-letter marker (mirrors vector/sharing_state.py). The scanner consults it
tenant-wide for every user and skips re-queuing until the content (etag) OR the
escalation-tier set (tiers_sig -- e.g. OCR enabled) changes, so the document is
attempted once per content-version instead of forever.
- new vector/dead_letter.py: mark/is/clear, content-addressed marker carrying
is_placeholder=True (inherits search exclusion) + dead_letter=True
- escalation.escalation_tiers_signature(settings): retry-on-tier-change key
- processor: dead-letter terminal failures, clear on successful (re-)index
- scanner: user-agnostic is_dead_lettered skip beside claim_existing_index
- placeholder: exempt dead_letter markers from the orphan sweep (durability)
- metrics: astrolabe_document_dead_lettered_total{reason}
Deck #349.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address round-4 review on PR #914:
- glyph_corruption_ratio <= 0 now disables the signal (previously `control_ratio
> 0` fired on any single C0 control byte), matching the "0 disables" convention
used elsewhere (document_max_pdf_size_mb) and the config comment. Add a
zero-disables test.
- Correct the document_escalation_suppressed_total comment: corrupt_glyphs CAN
appear there in the narrow case where structured is unregistered and OCR is
registered-but-disabled (evaluate_escalation follows minimum="structured" past
the missing rung to a gated-off OCR). Add a test for that suppressed decision.
- Add a test for the double-corruption edge: a structured re-extract that is also
glyph-corrupt escalates structured->ocr with reason corrupt_glyphs.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address round-2 review on PR #914:
- Add corrupt_glyphs to the document_classifier_flag_total label comment (it is
a live flag value emitted by record_document_classification).
- Mirror the full_text-vs-sampled control-ratio NOTE into classify_pdf so the
diagnostic path's under-detection trade-off is documented in place.
- Add test_classify_pdf_glyph_corrupt_routes_structured for routing symmetry on
the standalone classify_pdf path.
(SonarCloud quality gate is green — the prior S1244 finding was fixed last round.)
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address round-1 review on PR #914:
- Attribute the OCR hop in a fast->structured->ocr inline cascade to
from_tier="structured" (not a second "fast" escalation), so
astrolabe_document_escalation_total per-tier counts stay accurate.
- Add test_inline_fast_structured_ocr_cascade pinning that two-hop path and the
metric attribution.
- Note in classify_from_text that its doc-level control ratio is over full_text
(all pages), not the sampled subset classify_pdf uses.
- Clarify that corrupt_glyphs never lands in the suppressed-escalation counter.
- Dedupe the glyph-corrupt test string into tests/fixtures/glyph_corruption.py.
- Use pytest.approx for the control-char-ratio zero checks (SonarCloud S1244).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The fast (pypdfium2) extractor can leak raw glyph codes on subset fonts with a
broken /ToUnicode CMap. The result scores high on the existing text-quality
heuristic -- a uniform glyph/Caesar offset preserves whitespace and token
lengths -- yet is unsearchable. The structured (pymupdf) tier extracts the same
pages correctly.
Add a language-agnostic C0-control-character-ratio signal to the tier-0
classifier that detects this corruption and routes the document to a new
`structured` recommended_tier. Wire the fast->structured hop on the inline path
and generalise it so a low-quality-but-non-empty layer also tries structured
before OCR -- the inline and external ingest modes now follow the full
fast->structured->ocr ladder identically. A scanned / no-text-layer document
(total_chars == 0) still shortcuts straight to OCR, since a text extractor
cannot recover a pure raster.
New per-tenant tunable DOCUMENT_GLYPH_CORRUPTION_RATIO (default 0.02); escalation
metrics gain a `corrupt_glyphs` reason label.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
OCR is the paid, opt-in tier (DOCUMENT_OCR_ENABLED, default off). The per-tier
escalation gate already declines to hop to OCR when it's disabled (the pre-OCR
tier is terminal — no surprise cost), but that left operators blind to how much
OCR demand exists.
evaluate_escalation now returns a structured EscalationDecision:
- "hop" — a higher tier can run; the caller raises EscalateError (queue-hop).
- "suppressed" — the ideal next tier (e.g. ocr) exists but is DISABLED; the caller
indexes the current tier's output as terminal and records the
would-be hop on the new astrolabe_document_escalation_suppressed_total
{from_tier,to_tier,reason} counter instead of hopping.
- None — index as-is (good text, or no such tier at all).
So with OCR off, escalation_suppressed_total{to_tier="ocr"} is the latent OCR
demand an operator weighs before enabling OCR; enabling it converts these into
real document_escalation_total{to_tier="ocr"} hops. next_available_tier gains an
ignore_enabled flag to compute the *ideal* (enabled-gate-ignored) target.
Tests: registry suppressed vs hop vs terminal (incl. structured-hop-not-suppressed
when OCR off but structured available); _parse_pdf_tier records suppressed +
indexes without raising.
Deck #324 (parent #323).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- metrics: update_ingest_queue_depth guarded on `not by_queue`, which conflated
None (memory backend no-op) with {} (postgres, ALL queues drained). When every
queue drains at once, get_ingest_job_counts_by_queue returns {} and the
pre-zero loop was skipped, leaving a stale ghost backlog in the gauge. Guard on
`by_queue is None` only; add an all-drained regression test.
- procrastinate: note that INGEST_TRANSIENT_MAX_ATTEMPTS is snapshotted at
blueprint-build time (restart to pick up changes).
Deck #323.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- metrics: update_ingest_queue_depth now pre-zeroes every managed ingest queue
before applying live counts, so a queue that drains to empty (and drops out of
procrastinate's list_queues_async) reads 0 instead of sticking at its last
non-zero value (ghost backlog in Grafana/alerts). Adds a regression test.
- procrastinate: comment that _is_transient_infra_error treats all qdrant errors
as transient deliberately (bounded same-tier retry; over-broad is acceptable).
- escalation: note next_tier is the building block; production routing uses
ProcessorRegistry.next_available_tier.
- tests: add evaluate_escalation fast+ocr-only low-confidence -> ocr case.
Deck #323.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Split external (procrastinate) document processing into per-tier queues so a
document is attempted at most once per tier and requeued to the next tier's
queue on a low-quality parse, using procrastinate's native retry.
- escalation.py: TIER_LADDER (fast->structured->ocr) + EscalateError signal
- registry: process_tier (one tier) + evaluate_escalation post-parse gate
(reuses classify_from_text) + next_available_tier; shared _classify_result
and _oversize_result with the inline pipeline
- processor: process_document(tier=...) runs one tier and raises EscalateError
before embed (junk text never indexed); inline memory path unchanged
- queue/procrastinate: ingest-fast|structured|ocr queues; TieredEscalationStrategy
(queue-hop on EscalateError, bounded same-tier transient retry); queue-aware
task; producer defers to ingest-fast; per-queue counts + all-queue reclaim
- cli: worker --tier {fast,structured,ocr}
- billing: pages_ocr usage event + pipeline_tier metadata (paid OCR billed apart)
- observability: astrolabe_ingest_queue_depth{queue,status} gauge + per-queue
counts in nc_get_vector_sync_status / management status endpoint
- config: INGEST_ESCALATION_ENABLED (default true), INGEST_TRANSIENT_MAX_ATTEMPTS
INGEST_ESCALATION_ENABLED=false and INGEST_QUEUE=memory preserve prior behaviour.
Deck #323.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Round-3 review on PR #893:
- record_qdrant_operation("upsert","error") now fires only when the exhausted
retry was actually a Qdrant failure (reason=="qdrant"); an embed/connection
failure exhausts retries before Qdrant is called, so attributing it to
mcp_qdrant_operations_total{error} inflated that signal. The cause is still
captured by record_ingest_dropped.
- Add test_mistral_embed_retries_on_5xx: exercises the full Mistral retry path
(5xx SDKError then success), not just the predicate.
- Add test_generate_does_not_retry_on_bad_request: generate() fast-fails on a
permanent 4xx.
- Move astrolabe_vector_ingest_dropped_total's definition into the astrolabe_
pipeline-metrics block (was in the mcp_ section).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
From card 309 (OHR-Bench smoke-test triage): during a backend-pod rollover the
embedding endpoint was briefly unreachable, and openai.APIConnectionError /
ConnectError propagated unretried (the provider only retried 429). Documents
exhausted the 3 in-process retries and were dropped for that scan cycle.
Broaden the provider-level retry to the transient set -- APIConnectionError,
APITimeoutError, 429, and 5xx -- on the existing exponential backoff (2s->60s,
5 attempts), so a few seconds of retry rides through the rollover. Permanent
4xx (auth, bad request) still re-raise immediately. Generalize the shared
_retry helper (retry_on_rate_limit -> retry_on_transient, predicate renamed to
should_retry, accurate log label) with a back-compat alias; Mistral gets 429+5xx
for parity. The production gateway path inherits this via GatewayProvider, which
delegates to the decorated OpenAIProvider methods.
Add astrolabe_vector_ingest_dropped_total{reason}, incremented when a document
exhausts retries, classified (connection|timeout|rate_limit|server|qdrant|other)
by _drop_reason so the embed-drop rate is alertable per cause. Dropped docs are
NOT marked failed, so the next full scan re-picks them (re-queue via scan loop).
Refs: Deck board 12 card 309 (AC #1 no permanently-dropped docs; embed-drop
metric for AC #5).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Round-8 claude-review (no blockers; comment-only):
- 🟡 Documented that Ollama's /api/embed prompt_eval_count is assumed
batch-level total and is unverified against a live instance (Ollama isn't the
Cloud billing provider); if it proves last-item-only, switch to per-item
summing. The char estimate already covers versions that omit the field.
- 🟡 Noted on the astrolabe_embedding_tokens_total counter that operation="query"
is recorded pre-Qdrant, so it can legitimately exceed the billing-store
tokens_embedded aggregate when a search fails post-embed — dashboards
shouldn't alert on that healthy gap.
Deferred (reviewer: "minor nit, acceptable"): record_indexing_usage awaited in
the task group — the group awaits all child tasks regardless, the write is
best-effort + fast, and start_soon would need the tg threaded into the closure
for marginal gain.
Deck #284.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Billing product model finalized (Deck #281): bill pages externally, record
tokens internally. Rename the data-plane metric literals to match the now-
canonical contract (Deck #284) — the control plane's METRIC_EVENT_NAMES is
already renamed, so the old names would be unmapped and never sync to Stripe.
Rename (values unchanged):
- embeddings_queries → tokens_embedded (value = real token count, already
emitted by this PR; the unit upstream providers bill on).
- pages_chunks → pages_embedded (value kept as len(chunk_texts) interim;
TODO(#282): real normalized "pages indexed" count — real pages for paginated
types, chars/tokens-per-page constant otherwise — is deferred to the
instrumentation card, this only lands the name/contract).
- All literals, log strings, docstrings, comments, the migration comment, and
tests renamed; grep confirms zero old strings remain.
Observability (new): export embedding token cost to Prometheus as
astrolabe_embedding_tokens_total{provider,operation} (operation = index|query)
so the billed cost unit is visible in Grafana, not just the per-tenant billing
DB. Dedicated counter (doesn't inflate the existing chunk/request metrics) and
always-on (independent of USAGE_METERING_ENABLED, so OSS/self-host gets it).
Wired on both the indexing batch embed and the search query embed (query inside
the per-request cache-miss branch, so reused embeddings aren't double-counted).
Note: the rename orphans any pre-existing embeddings_queries/pages_chunks rows
in tenant app DBs (CP no longer maps them) — acceptable; pipeline is inert with
throwaway dev/sandbox data.
Deck #284 (folded into PR #875).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The hot-path classifier escalated to OCR purely on character count, so a
scanned/handwritten PDF with a low-quality embedded text layer (>16 chars/page
but garbled) routed `fast` and indexed the junk -- e.g. Student 147.pdf's
"Little Acoms Primary"/"0110912020", which pollutes the vector and demotes the
doc in search (Deck #207).
- classifier: recalibrate `_text_quality` with a long-token-fraction term that
detects word-merging (dropped inter-word spaces) -- the dominant junk-layer
failure the old whitespace/overlong(>20) terms missed. Measured: the Student
147 scan ~0.42 (60% pages junk) vs >=0.94 for clean digital docs.
- classify_from_text now routes on quality + scan: a page is OCR-worthy if
near-empty OR low text-quality OR (when OCR + scan detection are enabled) it's
mostly a raster image. New `image_coverage_per_page` re-opens the PDF for the
scan signal, so that cost is paid only by OCR-opted-in tenants. Thresholds are
passed in from per-tenant settings (keyword-only).
- config: 4 per-tenant settings -- DOCUMENT_OCR_MIN_TEXT_QUALITY (0.5),
DOCUMENT_OCR_PAGE_FRACTION (0.5), DOCUMENT_OCR_MIN_PAGE_CHARS (16),
DOCUMENT_OCR_DETECT_SCANNED (true) -- with range validators.
- metrics: new astrolabe_document_ocr_page_fraction histogram (the value the
page-fraction threshold acts on) alongside document_text_quality, so operators
can tune the OCR escalation per tenant (quality vs cost).
Escalation gate, OCR backends, and off-by-default behavior unchanged (#858).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Address PR #855 review (all non-blocking):
- classifier: _sample_indices now always includes the first AND last page (the
old evenly-spaced sample missed the tail, e.g. last sampled index 95 on a
100-page doc -- a scanned tail could be missed).
- classifier + metrics: document that flags are diagnostic and fire
independently of routing (image_heavy on ANY page vs the ocr route needing a
page FRACTION), so flag{image_heavy} is expected to exceed classified{ocr}.
- classifier: clarify the text-quality whitespace comment (caps at 12%) and note
the image double-count approximation (min() caps coverage).
- tests: add the scanned (no text layer) and bad_text_layer (junk text over an
image) flag paths, and a test pinning first/last-page sampling.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
First step of the tiered document-processor effort (Deck #203): a cheap, local
pre-pass that recommends which extraction tier a PDF should start in, emitting
metrics WITHOUT changing routing yet -- so we gather per-tenant doc-mix data
before turning escalation on.
document_processors/classifier.py: classify_pdf(content) -> DocClassification.
Page-sampled (bounded on large docs), <~1s. Cheap signals only -- text-layer
chars, a text-quality score (catches the "Student 147" failure where a text
layer exists but is mashed/space-less junk), and image coverage. A page that is
mostly a raster image routes to OCR: its content (handwriting, stamps) isn't in
any text layer. Deliberately no get_drawings/graphics-density signal -- it's
slow on the exact pages it'd flag, the hotfix's graphics_limit already makes the
parse safe, and the (future) tier-1 quality gate catches lost tables.
Validated on the sample corpus: born-digital 2-col arxiv and a digital student
record -> fast (tier 1); a scanned+handwritten form -> ocr (tier 3).
Wiring (vector/processor.py): _shadow_classify runs the classifier on PDFs in a
worker thread, best-effort (never blocks/fails indexing), gated by the new
DOCUMENT_CLASSIFY_ENABLED setting. Metrics: astrolabe_document_classified_total
{recommended_tier}, astrolabe_document_classifier_flag_total{flag},
astrolabe_document_text_quality histogram.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The document processor crash-looped on one pathological PDF: pymupdf4llm's
table/graphics detection over a page with ~1M vector path items ballooned past
the 2 GiB pod limit. The parse ran in a thread, so nothing could interrupt or
memory-bound it -- a single bad file OOM-killed the whole pod.
Run the parse in an isolated worker subprocess (anyio.to_process, cancellable)
with an RLIMIT_AS memory cap and a wall-clock timeout, so a pathological file
fails THAT document instead of the pod (new document_processors/_isolation.py).
Also pass graphics_limit (default 5000) to to_markdown -- validated to cut the
known trigger page from 112 s to 23 s with bounded memory.
On a permanent parse failure the processor returns success=False (instead of
raising, which would retry 3x); vector/processor.py marks the placeholder
"failed" and skips indexing, and the scanner stops re-queuing failed placeholders
until the file changes -- so a doomed file no longer churns.
New per-tenant (per-pod env) settings: DOCUMENT_PDF_GRAPHICS_LIMIT,
DOCUMENT_PARSE_TIMEOUT_SECONDS, DOCUMENT_PARSE_MEM_LIMIT_MB. New metric
astrolabe_document_parse_failed_total{reason=timeout|oom|error} surfaces hard
failures that previously killed the process before any except ran.
First PR of the tiered document-processor effort (Deck #199); tier 0/1/3
pipeline tracked separately.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The only queue metric, mcp_vector_sync_queue_size, was updated inline by the
single-user consumer (processor_task) but never by the multi-user consumer
(oauth_processor_task). On multi-user tenants (e.g. blackbox-demo, 5 users) the
gauge read 0 for 24h while the live anyio buffer held ~2214 pending documents
(shown by /api/v1/vector-sync/status). The "indexed" figure was also a chunk
count (16039 points ≈ 480 docs) mislabelled as documents.
Publish a consumer-independent snapshot from a periodic task
(vector/metrics_publisher.vector_sync_metrics_task), spawned in BOTH lifespan
task groups (single-user and multi-user) and every queue backend:
- mcp_vector_sync_pending_documents — outstanding work via
ingest_status.get_ingest_pending() (anyio buffer depth or procrastinate
todo+doing); also keeps the legacy queue_size gauge meaningful on all paths.
- mcp_vector_sync_indexed_documents — distinct documents, counted exactly and
cheaply via the one chunk_index=0 point per document (no facet).
- mcp_vector_sync_indexed_chunks — total non-placeholder points.
The /api/v1/vector-sync/status endpoint now returns indexed_documents (distinct
docs) AND indexed_chunks separately, so documents and chunks are no longer
conflated. The publisher uses approximate Qdrant counts (every-N-seconds gauge);
the on-demand endpoint counts exactly. New knob:
VECTOR_SYNC_METRICS_REFRESH_INTERVAL (default 20s). Fail-safe: a metrics refresh
never disturbs ingest.
BREAKING CHANGE: /api/v1/vector-sync/status field `indexed_documents` now holds
the distinct-document count (was the chunk count); the chunk count moved to the
new `indexed_chunks` field. The Astrolabe UI + the nc_get_vector_sync_status MCP
tool / userinfo page are harmonized in a follow-up (Deck #195).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The vector-sync scanner expanded a tagged folder into its PDF descendants via
`WebdavClient.find_by_type(scope=dir)` with no result limit. A WebDAV SEARCH
with no `<d:nresults>` returns only Nextcloud's default page (~100 on the
affected instance), so large tagged folders were silently truncated and most
documents were never queued for indexing (e.g. a 220-file folder yielded 100).
Add `search_files_all`, which pages the SEARCH to completion. It uses
`<d:firstresult>` offset paging where supported and, because Nextcloud 31
ignores offset (verified against a live instance), detects the repeated page
and falls back to a single bounded fetch with an explicit large `<d:nresults>`.
`find_all_by_type` wraps this and is now used for tagged-folder expansion;
`find_by_type` is unchanged for the interactive MCP tools.
Crossing `WEBDAV_SEARCH_MAX_RESULTS` logs a warning and increments the new
`astrolabe_document_scan_truncated_total` metric, so a coverage cap can never
again hide files silently.
Scope: this fixes discovery only. Cross-user double-processing of identical
shared files (point-ID collisions) is tracked separately.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Remaining items from the PR #831 Claude review:
- processor span symmetry: add "vector_sync.total_chars" to the sparse
embedding span (already on the dense span) and drop the redundant
"embedding.batch_size" attribute from both spans — it always equalled
vector_sync.chunk_count and would mislead once batching is split.
- metrics: document the deliberate "throughput counts only on full success"
contract in record_document_parse (partial extractions flagged
success=False are counted as a parse-error but never inflate
pages/chars/bytes throughput).
- config: extract _detect_base_provider() -> (family, model) as the single
source of truth for the provider-detection priority chain, shared by
get_embedding_model_name() and get_embedding_provider_family(). Preserves
the intentional gateway asymmetry (only the family method short-circuits).
- base.py: Optional[...] -> PEP 604 `... | None`; drop now-unused import.
Behavior unchanged (get_embedding_* outputs covered by test_config.py).
Refs Deck #175, PR #831.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Reviewer findings:
- Fix double-count of exhausted-retry failures: the inner final-retry branch
and the outer except both recorded a processing error. Consolidate to the
outer handler (single call site); inner branch keeps only the Qdrant-upsert
error metric. Regression test added.
- Deletes are no longer counted as indexing events: the delete success path
drops doc_type so astrolabe_documents_indexed_total is not inflated.
Regression test added.
- Reuse the already-resolved `settings` in _index_document instead of a second
get_settings() call.
- Use explicit `> 0` guards in record_document_parse / record_embedding instead
of truthiness checks.
SonarCloud:
- S1244 (BUG): replace float `==` equality in metric tests with pytest.approx.
- S5332 (hotspot): use https in the gateway-URL test fixture.
- S1192: extract the repeated "vector_sync.chunk_count" span-attribute literal
into a module constant.
Review nit: move the duplicated `_sample` test helper into a shared
`metric_sample` fixture in tests/unit/conftest.py.
Refs Deck #175, PR #831.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Make per-tier bottlenecks in the document-processing pipeline
(scan -> fetch -> parse -> chunk -> embed -> Qdrant upsert) visible via
metrics, traces, and structured logs. Today the document_processors layer
emits only a logger.info line: no metric, no span, and page counts live only
inside a log string. The single processing-duration histogram is unlabeled and
whole-document, so it cannot isolate parse vs embed vs upsert.
New astrolabe_* metric family (distinct from the mcp_* protocol metrics):
- astrolabe_document_parse_{duration_seconds,total} + pages/chars/bytes counters
recorded at the ProcessorRegistry.process() boundary (covers all current and
future processors uniformly)
- astrolabe_document_escalation_total (dormant; tiered-pipeline readiness)
- astrolabe_embedding_{duration_seconds,requests_total,chunks_total,chars_total}
- astrolabe_document_chunks_total, astrolabe_documents_indexed_total{source,status}
Tracing: new document_processor.parse child span + enriched embed/chunk span
attributes (provider/model/batch_size/chunk_count). Structured logs gain a
consistent field vocabulary (doc_id, doc_type, processor, tier, pages, chars,
byte_size, chunks, duration_ms, status) so Loki can aggregate without regex.
Tier-readiness: processor/tier are labels from day one and a tier property is
added to DocumentProcessor, so adding docling/OCR/LLM tiers later is additive
(new label values, never new metrics). Tenant comes from the kube namespace
label; mime_type/model are span attributes only (cardinality). Existing
mcp_vector_sync_*/mcp_qdrant_* are left untouched.
Refs Deck #175 (superset of #173 Phase 2). Dashboard/recording-rules follow-up
tracked on #175 for homelab-argocd.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Many identity providers (AWS Cognito, Okta, Azure AD) reject or mishandle
colons in OAuth scope names. This migrates all custom scopes from
`resource:action` to `resource.action` format (e.g., `notes:read` →
`notes.read`), which is universally accepted and aligns with industry
conventions (Microsoft, Google).
Includes Alembic migration 004 for stored scope strings and ADR-024
documenting the rationale and RFC references.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Enable ruff PLC0415 rule for all source files (tests excluded via
per-file-ignores). Move 136 inline imports to top-level across 33 files.
8 imports suppressed with noqa for legitimate reasons: circular
dependencies (client/__init__.py, context.py), optional dependency
guards (app.py document processors, auth/userinfo_routes.py), and
post-env-setup imports (smithery_main.py).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Created @instrument_tool decorator for automatic MCP tool metrics collection.
Applied to all 7 tools in notes.py.
Changes:
- observability/metrics.py:
* New instrument_tool() decorator for automatic timing and error tracking
* Compatible with @mcp.tool() and @require_scopes() decorators
* Records tool_name, duration, and success/error status
- server/notes.py:
* Applied @instrument_tool to all 7 tool functions
* nc_notes_create_note, nc_notes_update_note, nc_notes_append_content
* nc_notes_search_notes, nc_notes_get_note, nc_notes_get_attachment
* nc_notes_delete_note
These metrics will populate the MCP Tool Calls dashboard panels.
Part of PR #295 - Complete metrics instrumentation (Phase 5)
Remaining: 86 tools across 8 server files
Security fix: Move Prometheus metrics endpoint from main HTTP port to
dedicated port 9090 to prevent external exposure of metrics data.
Changes:
- Use prometheus_client.start_http_server() for dedicated metrics server
- Remove /metrics route from main application routes
- Metrics now only accessible on port 9090 (configurable via METRICS_PORT)
- Main application port no longer serves /metrics endpoint
This follows security best practice of isolating monitoring endpoints
from application traffic.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Add Prometheus metrics for HTTP, MCP tools, Nextcloud API, OAuth, vector sync, and DB operations
- Add OpenTelemetry distributed tracing with OTLP export
- Add structured JSON logging with trace context correlation
- Add ObservabilityMiddleware for automatic HTTP instrumentation
- Add app_name attribute to all client classes for per-app metrics
- Add configuration for metrics, tracing, and logging via environment variables
- Add documentation in docs/observability.md
- Fix graceful degradation when tracing is disabled (default state)
- Fix uvicorn logging configuration to use observability formatters
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
