From 065502f562f493b8984f2ade86871638ee4df16e Mon Sep 17 00:00:00 2001
From: Jonathan Schweder <jonathanschweder@gmail.com>
Date: Tue, 16 Dec 2025 15:34:50 +0000
Subject: [PATCH] Create idx_bucket_data_compact index

---
 ...1734384000000-bucket-data-compact-index.ts | 41 +++++++++++++++++++
 .../src/storage/PostgresCompactor.ts          |  6 +--
 .../src/__snapshots__/storage.test.ts.snap    |  2 +-
 3 files changed, 45 insertions(+), 4 deletions(-)
 create mode 100644 modules/module-postgres-storage/src/migrations/scripts/1734384000000-bucket-data-compact-index.ts

diff --git a/modules/module-postgres-storage/src/migrations/scripts/1734384000000-bucket-data-compact-index.ts b/modules/module-postgres-storage/src/migrations/scripts/1734384000000-bucket-data-compact-index.ts
new file mode 100644
index 000000000..1a695ea42
--- /dev/null
+++ b/modules/module-postgres-storage/src/migrations/scripts/1734384000000-bucket-data-compact-index.ts
@@ -0,0 +1,41 @@
+import { migrations } from '@powersync/service-core';
+
+import { openMigrationDB } from '../migration-utils.js';
+
+/**
+ * Migration: Add specialized index for bucket_data compaction query optimization
+ *
+ * This migration creates an index optimized for the compaction query in PostgresCompactor.ts
+ * which uses COLLATE "C" (binary collation) and DESC ordering.
+ * The index enables PostgreSQL to push COLLATE "C" conditions into Index Cond
+ * instead of applying them as filters, significantly reducing rows scanned.
+ */
+export const up: migrations.PowerSyncMigrationFunction = async (context) => {
+  const {
+    service_context: { configuration }
+  } = context;
+  await using client = openMigrationDB(configuration.storage);
+
+  await client.transaction(async (db) => {
+    await db.sql`
+      CREATE INDEX IF NOT EXISTS idx_bucket_data_compact ON bucket_data (
+        group_id,
+        bucket_name COLLATE "C" DESC,
+        op_id DESC
+      )
+    `.execute();
+
+    await db.sql`ANALYZE bucket_data`.execute();
+  });
+};
+
+export const down: migrations.PowerSyncMigrationFunction = async (context) => {
+  const {
+    service_context: { configuration }
+  } = context;
+  await using client = openMigrationDB(configuration.storage);
+
+  await client.transaction(async (db) => {
+    await db.sql`DROP INDEX IF EXISTS idx_bucket_data_compact`.execute();
+  });
+};
diff --git a/modules/module-postgres-storage/src/storage/PostgresCompactor.ts b/modules/module-postgres-storage/src/storage/PostgresCompactor.ts
index b19d866eb..3d5f89185 100644
--- a/modules/module-postgres-storage/src/storage/PostgresCompactor.ts
+++ b/modules/module-postgres-storage/src/storage/PostgresCompactor.ts
@@ -123,16 +123,16 @@ export class PostgresCompactor {
           bucket_data
         WHERE
           group_id = ${{ type: 'int4', value: this.group_id }}
-          AND bucket_name >= ${{ type: 'varchar', value: bucketLower }}
+          AND bucket_name >= ${{ type: 'varchar', value: bucketLower }} COLLATE "C"
           AND (
             (
-              bucket_name = ${{ type: 'varchar', value: bucketUpper }}
+              bucket_name = ${{ type: 'varchar', value: bucketUpper }} COLLATE "C"
               AND op_id < ${{ type: 'int8', value: upperOpIdLimit }}
             )
             OR bucket_name < ${{ type: 'varchar', value: bucketUpper }} COLLATE "C" -- Use binary comparison
           )
         ORDER BY
-          bucket_name DESC,
+          bucket_name COLLATE "C" DESC,
           op_id DESC
         LIMIT
           ${{ type: 'int4', value: this.moveBatchQueryLimit }}
diff --git a/modules/module-postgres-storage/test/src/__snapshots__/storage.test.ts.snap b/modules/module-postgres-storage/test/src/__snapshots__/storage.test.ts.snap
index d1b24f45b..79e1e6242 100644
--- a/modules/module-postgres-storage/test/src/__snapshots__/storage.test.ts.snap
+++ b/modules/module-postgres-storage/test/src/__snapshots__/storage.test.ts.snap
@@ -2,7 +2,7 @@
 
 exports[`Postgres Sync Bucket Storage - Data > empty storage metrics 1`] = `
 {
-  "operations_size_bytes": 16384,
+  "operations_size_bytes": 24576,
   "parameters_size_bytes": 32768,
   "replication_size_bytes": 16384,
 }