release_chunks' indexes with its chunks (CHUNK_TABLESPACE)

The October partition's three indexes took 2.9 GB of the NVMe, its pkey
never scanned; chunks are read from the HDD anyway. Moved on homelab by hand
(12 s); the migration does the same on a new deploy, parent indexes too so new
partitions follow.

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

database/migrations/1770000003100_chunk_indexes_with_chunks.ts

@@ -0,0 +1,26 @@
+import { BaseSchema } from '@adonisjs/lucid/schema'
+import env from '#start/env'
+
+/**
+ * release_chunks' indexes with its chunks in CHUNK_TABLESPACE (the HDD), not the catalog's tablespace: a chunk is
+ * read from the HDD anyway, and the October partition's three indexes took 2.9 GB of the NVMe (its pkey never
+ * scanned; LOG.md 2026-10-14 (54)). The parent's partitioned indexes too, so new partitions put theirs there.
+ * Homelab was moved by hand first; a no-op where they're there already.
+ */
+export default class extends BaseSchema {
+  async up() {
+    const tablespace = env.get('CHUNK_TABLESPACE')
+    if (!tablespace) return
+    this.defer(async (db) => {
+      const { rows } = await db.rawQuery(
+        `SELECT i.relname FROM pg_class c JOIN pg_index x ON x.indrelid = c.oid JOIN pg_class i ON i.oid = x.indexrelid
+         LEFT JOIN pg_tablespace t ON t.oid = i.reltablespace
+         WHERE c.relname LIKE 'release_chunks%' AND coalesce(t.spcname, 'pg_default') <> ?`,
+        [tablespace]
+      )
+      for (const { relname } of rows) await db.rawQuery(`ALTER INDEX ${relname} SET TABLESPACE ${tablespace}`)
+    })
+  }
+
+  async down() {}
+}