Improve performance of "phabricator:20210215.changeset.02.phid-populate.php"

Summary: Ref T13613. Improve the performance of this migration by using a temporary table and an "UPDATE x JOIN y ..." pattern.

Test Plan:
  - Ran on `secure`, got exit after a few seconds since the migration is idempotent and changesets already had PHIDs.
  - Ran on `secure` with the `continue;` commented out, got valid new PHIDs in 53s (from 153s).
  - Tried a larger page size (16K), didn't see any improvement.
  - From "--trace", client PHID generation seems to be the limiting factor.

Subscribers: PHID-OPKG-gm6ozazyms6q6i22gyam

Maniphest Tasks: T13613

Differential Revision: https://secure.phabricator.com/D21570

Author: epriestley

resources/sql/autopatches/20210215.changeset.02.phid-populate.php

@@ -7,20 +7,65 @@
 $conn = $changeset_table->establishConnection('w');
 $table_name = $changeset_table->getTableName();
 
-$iterator = new LiskRawMigrationIterator($conn, $table_name);
-foreach ($iterator as $changeset_row) {
-  $phid = $changeset_row['phid'];
+$chunk_size = 4096;
 
-  if (strlen($phid)) {
+$temporary_table = 'tmp_20210215_changeset_id_map';
+
+queryfx(
+  $conn,
+  'CREATE TEMPORARY TABLE %T (
+    changeset_id INT UNSIGNED NOT NULL AUTO_INCREMENT PRIMARY KEY,
+    changeset_phid VARBINARY(64) NOT NULL)',
+  $temporary_table);
+
+$table_iterator = id(new LiskRawMigrationIterator($conn, $table_name))
+  ->setPageSize($chunk_size);
+
+$chunk_iterator = new PhutilChunkedIterator($table_iterator, $chunk_size);
+foreach ($chunk_iterator as $chunk) {
+
+  $map = array();
+  foreach ($chunk as $changeset_row) {
+    $phid = $changeset_row['phid'];
+
+    if (strlen($phid)) {
+      continue;
+    }
+
+    $phid = PhabricatorPHID::generateNewPHID($phid_type);
+    $id = $changeset_row['id'];
+
+    $map[(int)$id] = $phid;
+  }
+
+  if (!$map) {
     continue;
   }
 
-  $phid = PhabricatorPHID::generateNewPHID($phid_type);
+  $sql = array();
+  foreach ($map as $changeset_id => $changeset_phid) {
+    $sql[] = qsprintf(
+      $conn,
+      '(%d, %s)',
+      $changeset_id,
+      $changeset_phid);
+  }
+
+  queryfx(
+    $conn,
+    'TRUNCATE TABLE %T',
+    $temporary_table);
+
+  queryfx(
+    $conn,
+    'INSERT INTO %T (changeset_id, changeset_phid) VALUES %LQ',
+    $temporary_table,
+    $sql);
 
   queryfx(
     $conn,
-    'UPDATE %T SET phid = %s WHERE id = %d',
+    'UPDATE %T c JOIN %T x ON c.id = x.changeset_id
+      SET c.phid = x.changeset_phid',
     $table_name,
-    $phid,
-    $changeset_row['id']);
+    $temporary_table);
 }