From 4dc5cd610eef22e128058bf85d6457727064af8c Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?J=C3=A9r=C3=B4me=20Tamarelle?= <jerome@tamarelle.net>
Date: Tue, 19 Sep 2023 22:07:55 +0200
Subject: [PATCH] PHPLIB-1237 Implement parallel benchmarks

---
 benchmark/DriverBench/ParallelBench.php | 193 ++++++++++++++++++++++++
 benchmark/Fixtures/Data.php             |   1 +
 benchmark/Fixtures/data/ldjson.json     |   1 +
 benchmark/Utils.php                     |   7 +
 4 files changed, 202 insertions(+)
 create mode 100644 benchmark/DriverBench/ParallelBench.php
 create mode 100644 benchmark/Fixtures/data/ldjson.json

diff --git a/benchmark/DriverBench/ParallelBench.php b/benchmark/DriverBench/ParallelBench.php
new file mode 100644
index 000000000..e55af9f0d
--- /dev/null
+++ b/benchmark/DriverBench/ParallelBench.php
@@ -0,0 +1,193 @@
+<?php
+
+namespace MongoDB\Benchmark\DriverBench;
+
+use Generator;
+use MongoDB\Benchmark\Fixtures\Data;
+use MongoDB\Benchmark\Utils;
+use MongoDB\BSON\Document;
+use MongoDB\Collection;
+use PhpBench\Attributes\AfterClassMethods;
+use PhpBench\Attributes\BeforeClassMethods;
+use PhpBench\Attributes\BeforeMethods;
+use PhpBench\Attributes\ParamProviders;
+use PhpBench\Attributes\Revs;
+use RuntimeException;
+
+use function array_chunk;
+use function array_map;
+use function ceil;
+use function count;
+use function file;
+use function file_get_contents;
+use function file_put_contents;
+use function is_dir;
+use function mkdir;
+use function pcntl_fork;
+use function pcntl_waitpid;
+use function range;
+use function sprintf;
+use function str_repeat;
+use function sys_get_temp_dir;
+use function unlink;
+
+use const FILE_IGNORE_NEW_LINES;
+use const FILE_NO_DEFAULT_CONTEXT;
+use const FILE_SKIP_EMPTY_LINES;
+
+/**
+ * For accurate results, run benchmarks on a standalone server.
+ *
+ * @see https://github.com/mongodb/specifications/blob/ddfc8b583d49aaf8c4c19fa01255afb66b36b92e/source/benchmarking/benchmarking.rst#parallel
+ */
+#[BeforeClassMethods('beforeClass')]
+#[AfterClassMethods('afterClass')]
+final class ParallelBench
+{
+    /** @var string[] */
+    private static array $files = [];
+
+    public static function beforeClass(): void
+    {
+        // Generate files
+        $fileContents = str_repeat(file_get_contents(Data::LDJSON_FILE_PATH), 5_000);
+        foreach (self::getFileNames() as $file) {
+            file_put_contents($file, $fileContents);
+        }
+    }
+
+    public static function afterClass(): void
+    {
+        foreach (self::getFileNames() as $file) {
+            unlink($file);
+        }
+
+        self::$files = [];
+    }
+
+    /**
+     * Parallel: LDJSON multi-file import
+     * Using single thread
+     *
+     * @see https://github.com/mongodb/specifications/blob/ddfc8b583d49aaf8c4c19fa01255afb66b36b92e/source/benchmarking/benchmarking.rst#ldjson-multi-file-import
+     */
+    #[BeforeMethods('beforeMultiFileImport')]
+    #[Revs(1)]
+    public function benchMultiFileImport(): void
+    {
+        $collection = Utils::getCollection();
+        foreach (self::getFileNames() as $file) {
+            self::importFile($file, $collection);
+        }
+    }
+
+    /**
+     * Parallel: LDJSON multi-file import
+     * Using multiple forked threads
+     *
+     * @see https://github.com/mongodb/specifications/blob/ddfc8b583d49aaf8c4c19fa01255afb66b36b92e/source/benchmarking/benchmarking.rst#ldjson-multi-file-import
+     * @param array{processes:int, files:string[], batchSize:int} $params
+     */
+    #[BeforeMethods('beforeMultiFileImport')]
+    #[ParamProviders(['provideProcessesParameter', 'provideMultiFileImportParameters'])]
+    #[Revs(1)]
+    public function benchMultiFileImportFork(array $params): void
+    {
+        $pids = [];
+        foreach ($params['files'] as $files) {
+            // Wait for a child process to finish if we have reached the maximum number of processes
+            if (count($pids) >= $params['processes']) {
+                $pid = pcntl_waitpid(-1, $status);
+                unset($pids[$pid]);
+            }
+
+            $pid = pcntl_fork();
+            if ($pid === 0) {
+                // If we reset, we can garantee that we get a new manager in the child process
+                // If we don't reset, we will get the same manager client_zval in the child process
+                // and share the libmongoc client.
+                Utils::reset();
+                $collection = Utils::getCollection();
+
+                foreach ($files as $file) {
+                    self::importFile($file, $collection);
+                }
+
+                // Exit the child process
+                exit(0);
+            }
+
+            if ($pid === -1) {
+                throw new RuntimeException('Failed to fork');
+            }
+
+            // Keep the forked process id to wait for it later
+            $pids[$pid] = true;
+        }
+
+        // Wait for all child processes to finish
+        while ($pids !== []) {
+            $pid = pcntl_waitpid(-1, $status);
+            unset($pids[$pid]);
+        }
+    }
+
+    public static function provideProcessesParameter(): Generator
+    {
+        // Max number of forked processes
+        for ($i = 1; $i <= 30; $i = (int) ceil($i * 1.25)) {
+            yield $i . 'fork' => ['processes' => $i];
+        }
+    }
+
+    public static function provideMultiFileImportParameters(): Generator
+    {
+        $files = self::getFileNames();
+
+        // Chunk of file names to be handled by each processes
+        for ($i = 1; $i <= 10; $i += 3) {
+            yield 'by ' . $i => ['files' => array_chunk($files, $i)];
+        }
+    }
+
+    public function beforeMultiFileImport(): void
+    {
+        $database = Utils::getDatabase();
+        $database->drop();
+        $database->createCollection(Utils::getCollectionName());
+    }
+
+    public function afterMultiFileImport(): void
+    {
+        foreach (self::$files as $file) {
+            unlink($file);
+        }
+
+        unset($this->files);
+    }
+
+    private static function importFile(string $file, Collection $collection): void
+    {
+        // Read file contents into BSON documents
+        $docs = array_map(
+            static fn (string $line) => Document::fromJSON($line),
+            file($file, FILE_IGNORE_NEW_LINES | FILE_SKIP_EMPTY_LINES | FILE_NO_DEFAULT_CONTEXT),
+        );
+        // Insert documents in bulk
+        $collection->insertMany($docs);
+    }
+
+    private static function getFileNames(): array
+    {
+        $tempDir = sys_get_temp_dir() . '/mongodb-php-benchmark';
+        if (! is_dir($tempDir)) {
+            mkdir($tempDir);
+        }
+
+        return array_map(
+            static fn (int $i) => sprintf('%s/%03d.txt', $tempDir, $i),
+            //range(0, 99),
+            range(0, 5),
+        );
+    }
+}
diff --git a/benchmark/Fixtures/Data.php b/benchmark/Fixtures/Data.php
index a60d3d571..9c21771d4 100644
--- a/benchmark/Fixtures/Data.php
+++ b/benchmark/Fixtures/Data.php
@@ -16,6 +16,7 @@ final class Data
     public const LARGE_FILE_PATH = __DIR__ . '/data/large_doc.json';
     public const SMALL_FILE_PATH = __DIR__ . '/data/small_doc.json';
     public const TWEET_FILE_PATH = __DIR__ . '/data/tweet.json';
+    public const LDJSON_FILE_PATH = __DIR__ . '/data/ldjson.json';
 
     public static function readJsonFile(string $path): array
     {
diff --git a/benchmark/Fixtures/data/ldjson.json b/benchmark/Fixtures/data/ldjson.json
new file mode 100644
index 000000000..3b1421232
--- /dev/null
+++ b/benchmark/Fixtures/data/ldjson.json
@@ -0,0 +1 @@
+{"text":"@wildfits you're not getting one.....","in_reply_to_status_id":22773233453,"retweet_count":null,"contributors":null,"created_at":"Thu Sep 02 19:38:18 +0000 2010","geo":null,"source":"web","coordinates":null,"in_reply_to_screen_name":"wildfits","truncated":false,"entities":{"user_mentions":[{"indices":[0,9],"screen_name":"wildfits","name":"Mairin Goetzinger","id":41832464}],"urls":[],"hashtags":[]},"retweeted":false,"place":null,"user":{"friends_count":179,"profile_sidebar_fill_color":"7a7a7a","location":"Minneapols, MN/Brookings SD","verified":false,"follow_request_sent":null,"favourites_count":0,"profile_sidebar_border_color":"a3a3a3","profile_image_url":"http://a1.twimg.com/profile_images/1110614677/Screen_shot_2010-08-25_at_10.12.40_AM_normal.png","geo_enabled":false,"created_at":"Sun Aug 17 00:23:13 +0000 2008","description":"graphic designer + foodie, with a love of music, movies, running, design, + the outdoors!","time_zone":"Mountain Time (US & Canada)","url":"http://jessiefarris.com/","screen_name":"jessiekf","notifications":null,"profile_background_color":"303030","listed_count":1,"lang":"en"}}
diff --git a/benchmark/Utils.php b/benchmark/Utils.php
index c19221be4..43ac051d9 100644
--- a/benchmark/Utils.php
+++ b/benchmark/Utils.php
@@ -43,4 +43,11 @@ public static function getCollectionName(): string
     {
         return 'perftest';
     }
+
+    public static function reset(): void
+    {
+        self::$client = null;
+        self::$database = null;
+        self::$collection = null;
+    }
 }