Inspection backlog in Postgres; BullMQ gets a working set

5.4M waiting inspection jobs took 3.3 GB of Redis and most of its single
thread. Producers now add file ids to inspection_queue; index-scheduler
feeds up to INSPECTION_FEED_MAX (50k) into BullMQ. inspection:from-redis
moves the existing backlog over once.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

app/services/BinaryFileService.ts

@@ -1,8 +1,8 @@
 import db from '@adonisjs/lucid/services/db'
 import FileStore from '#services/FileStore'
 import SfvService from '#services/SfvService'
-import QueueService from '#services/QueueService'
 import ContentService from '#services/ContentService'
+import InspectionQueue from '#services/InspectionQueue'
 
 export interface BinaryFilesResult {
   /** New files. */
@@ -134,8 +134,8 @@ class BinaryFileService {
     // By the posted name, as before: inspection skips later volumes and readable media names anyway.
     const jobs = added
       .filter((file) => ContentService.isCandidate(file.posted))
-      .map((file) => ({ name: 'process-collection', data: { fileId: file.id }, opts: { jobId: `file-${file.id}` } }))
-    if (jobs.length > 0) await QueueService.collectionQueue.addBulk(jobs)
+      .map((file) => file.id)
+    if (jobs.length > 0) await InspectionQueue.add(jobs)
     result.inspections = jobs.length
     return result
   }