88da265ef4
Measured during a live album download: ~4m15s per track in post-processing (normal is ~20s), with the time vanishing silently inside embed_source_ids — up to 5 MusicBrainz calls per track crawling against a degraded musicbrainz.org while the MB enrichment worker kept eating the same ~1 req/s per-IP budget. Only Spotify/Last.fm/Genius were in the yield set; MusicBrainz, Deezer, iTunes, Discogs etc. kept grinding through downloads. Policy (new core/enrichment/yield_policy, tested): - downloads active -> ALL enrichment workers yield (post-processing touches every metadata source). listening-stats (local-only) and repair (user-scheduled) intentionally keep running. - discovery active -> the API-contention five yield (spotify/itunes/deezer/ discogs/hydrabase) — discovery never paused anything before, despite the pause helper literally defaulting to label='discovery'. - user overrides and user-paused bookkeeping keep their existing semantics; the dashboard yield_reason label now says WHICH foreground work caused it. Observability (the 4-minute silence can never come back): - every source lookup is timed; >2s logs a warning NAMING the source and duration (core/metadata/source.py _call_source_lookup) - the pipeline always logs "Metadata enhancement took X.Xs" per track 7 policy tests (incl. the motivating case: MB yields to downloads, keeps running during discovery); 277 pipeline/enrichment tests pass.
60 lines
2.3 KiB
Python
60 lines
2.3 KiB
Python
"""Tests for the enrichment-worker yield policy (downloads/discovery contention)."""
|
|
|
|
from __future__ import annotations
|
|
|
|
from core.enrichment.yield_policy import (
|
|
ALL_YIELD_WORKERS,
|
|
API_CONTENTION_WORKERS,
|
|
discovery_state_active,
|
|
worker_yield_reason,
|
|
)
|
|
|
|
|
|
def test_downloads_pause_everything():
|
|
for name in ALL_YIELD_WORKERS:
|
|
assert worker_yield_reason(name, downloading=True, discovering=False) == 'downloads'
|
|
|
|
|
|
def test_discovery_pauses_only_the_contention_five():
|
|
for name in ALL_YIELD_WORKERS:
|
|
reason = worker_yield_reason(name, downloading=False, discovering=True)
|
|
if name in API_CONTENTION_WORKERS:
|
|
assert reason == 'discovery', name
|
|
else:
|
|
assert reason is None, name
|
|
|
|
|
|
def test_downloads_outrank_discovery_for_the_label():
|
|
assert worker_yield_reason('spotify-enrichment', True, True) == 'downloads'
|
|
|
|
|
|
def test_idle_pauses_nothing():
|
|
for name in ALL_YIELD_WORKERS:
|
|
assert worker_yield_reason(name, False, False) is None
|
|
|
|
|
|
def test_unknown_and_excluded_workers_never_yield():
|
|
# listening-stats (local media server only) and repair (user-scheduled job
|
|
# runner) intentionally keep running through downloads.
|
|
for name in ('listening-stats', 'repair', 'definitely-not-a-worker'):
|
|
assert worker_yield_reason(name, True, True) is None
|
|
|
|
|
|
def test_musicbrainz_yields_for_downloads_not_discovery():
|
|
# The case that motivated all of this: the MB worker starving the import
|
|
# pipeline's per-track lookups (~4m15s/track measured). It must yield to
|
|
# downloads — but keep running during discovery, which doesn't use MB.
|
|
assert worker_yield_reason('musicbrainz', True, False) == 'downloads'
|
|
assert worker_yield_reason('musicbrainz', False, True) is None
|
|
|
|
|
|
def test_discovery_state_active_phases():
|
|
assert discovery_state_active({'phase': 'discovering'})
|
|
assert discovery_state_active({'phase': 'Matching tracks...'})
|
|
assert not discovery_state_active({'phase': 'idle'})
|
|
assert not discovery_state_active({'phase': ''})
|
|
assert not discovery_state_active({'phase': 'discovered'})
|
|
assert not discovery_state_active({'phase': 'error'})
|
|
assert not discovery_state_active({'phase': 'cancelled'})
|
|
assert not discovery_state_active({})
|
|
assert not discovery_state_active(None)
|