Ship production optimization M1-M12.5A: queues, metrics, HTTP observability, and vector search reliability.
Keep similar-ai from tripping the global circuit on a lone URL 502, clamp Qdrant search to 100, and add Server-Timing plus slow-request logging. Studio shared props, Academy S3 exists caching, heat chunking, and Redis/scheduler hygiene stay in this rollout.
This commit is contained in:
@@ -77,6 +77,11 @@ return [
|
||||
'candidate_pool' => 200, // top N candidates to run diversity filter on
|
||||
],
|
||||
|
||||
// Unique lock for RankBuildScopeListsJob. Must exceed queue wait + job
|
||||
// timeout (300s). Production wait has been ~2.5h; 6h prevents hourly
|
||||
// stacking while still recovering a dead lock the same day.
|
||||
'scope_job_unique_for' => (int) env('RANK_SCOPE_JOB_UNIQUE_FOR', 21600),
|
||||
|
||||
// ── Anti-spam / burst-view damping ──────────────────────────────────────
|
||||
'spam' => [
|
||||
'views_24h_threshold' => 2000,
|
||||
|
||||
Reference in New Issue
Block a user