chore: Adjust benchmark runs to give more predictable results

phpbench.json.dist

@@ -31,7 +31,7 @@
   "runner.php_env": {
     "PGSQL_DATABASE_URL": "pgsql://postgres:[email protected]:5432/postgres?serverVersion=11&charset=utf8"
   },
-  "runner.iterations": 2,
-  "runner.retry_threshold": 10,
+  "runner.iterations": 10,
+  "runner.retry_threshold": 5,
   "storage.xml_storage_path": "var/phpbench"
 }

src/adapter/etl-adapter-csv/tests/Flow/ETL/Adapter/CSV/Tests/Benchmark/CSVLoaderBench.php

@@ -22,7 +22,7 @@ public function __construct()
     {
         $this->context = flow_context(config());
         $this->outputPath = \tempnam(\sys_get_temp_dir(), 'etl_csv_loader_bench') . '.csv';
-        $this->rows = (new FakeStaticOrdersExtractor(10_000))->toRows();
+        $this->rows = (new FakeStaticOrdersExtractor(1_000))->toRows();
     }
 
     public function __destruct()
@@ -34,7 +34,7 @@ public function __destruct()
         \unlink($this->outputPath);
     }
 
-    public function bench_load_10k() : void
+    public function bench_load_1k() : void
     {
         to_csv($this->outputPath)->load($this->rows, $this->context);
     }

src/adapter/etl-adapter-doctrine/tests/Flow/ETL/Adapter/Doctrine/Tests/Benchmark/DbalExtractorBench.php

@@ -6,25 +6,21 @@
 
 use function Flow\ETL\Adapter\Doctrine\{from_dbal_key_set_qb, from_dbal_limit_offset_qb, pagination_key_asc, pagination_key_set, to_dbal_schema_table, to_dbal_table_insert};
 use function Flow\ETL\DSL\{config, df, flow_context};
-use Doctrine\DBAL\{DriverManager, ParameterType};
+use Doctrine\DBAL\{Connection, DriverManager, ParameterType};
 use Doctrine\DBAL\Tools\DsnParser;
 use Flow\ETL\Tests\Double\FakeStaticOrdersExtractor;
 use PhpBench\Attributes\Groups;
 
 #[Groups(['extractor'])]
 final class DbalExtractorBench
 {
-    private const TABLE_NAME = 'benchmark_orders_extractor';
+    private const string TABLE_NAME = 'benchmark_orders_extractor';
 
-    private \Doctrine\DBAL\Connection $connection;
+    private Connection $connection;
 
     public function __construct()
     {
-        $dsn = \getenv('PGSQL_DATABASE_URL');
-
-        if (!$dsn) {
-            throw new \RuntimeException('PGSQL_DATABASE_URL environment variable is not set');
-        }
+        $dsn = \getenv('PGSQL_DATABASE_URL') ?: throw new \RuntimeException('PGSQL_DATABASE_URL environment variable is not set');
 
         $params = (new DsnParser(['postgresql' => 'pdo_pgsql']))->parse($dsn);
 
@@ -44,7 +40,7 @@ public function __destruct()
         $this->connection->close();
     }
 
-    public function bench_extract_10k_keyset() : void
+    public function bench_extract_1k_keyset() : void
     {
         $context = flow_context(config());
 
@@ -60,7 +56,7 @@ public function bench_extract_10k_keyset() : void
         }
     }
 
-    public function bench_extract_10k_limit_offset() : void
+    public function bench_extract_1k_limit_offset() : void
     {
         $context = flow_context(config());
 
@@ -72,7 +68,6 @@ public function bench_extract_10k_limit_offset() : void
         foreach (from_dbal_limit_offset_qb(
             $this->connection,
             $queryBuilder,
-            page_size: 1000
         )->extract($context) as $rows) {
         }
     }
@@ -89,7 +84,7 @@ private function setupDatabase() : void
         $table->setPrimaryKey(['index']);
         $schemaManager->createTable($table);
 
-        $extractor = new FakeStaticOrdersExtractor(10_000);
+        $extractor = new FakeStaticOrdersExtractor(1_000);
 
         df()
             ->read($extractor)

src/adapter/etl-adapter-doctrine/tests/Flow/ETL/Adapter/Doctrine/Tests/Benchmark/DbalLoaderBench.php

@@ -15,7 +15,7 @@
 #[Groups(['loader'])]
 final class DbalLoaderBench
 {
-    private const TABLE_NAME = 'benchmark_orders_loader';
+    private const string TABLE_NAME = 'benchmark_orders_loader';
 
     private Connection $connection;
 
@@ -25,16 +25,12 @@ final class DbalLoaderBench
 
     public function __construct()
     {
-        $dsn = \getenv('PGSQL_DATABASE_URL');
-
-        if (!$dsn) {
-            throw new \RuntimeException('PGSQL_DATABASE_URL environment variable is not set');
-        }
+        $dsn = \getenv('PGSQL_DATABASE_URL') ?: throw new \RuntimeException('PGSQL_DATABASE_URL environment variable is not set');
 
         $params = (new DsnParser(['postgresql' => 'pdo_pgsql']))->parse($dsn);
 
         $this->connection = DriverManager::getConnection($params);
-        $this->rows = (new FakeStaticOrdersExtractor(10_000))->toRows();
+        $this->rows = (new FakeStaticOrdersExtractor(1_000))->toRows();
         $this->context = flow_context();
     }
 
@@ -63,9 +59,9 @@ public function setUp() : void
     }
 
     #[BeforeMethods('setUp')]
-    public function bench_load_10k() : void
+    public function bench_load_1k() : void
     {
-        foreach ($this->rows->chunks(1_000) as $chunk) {
+        foreach ($this->rows->chunks(100) as $chunk) {
             to_dbal_table_insert($this->connection, self::TABLE_NAME)->load($chunk, $this->context);
         }
     }

src/adapter/etl-adapter-json/tests/Flow/ETL/Adapter/JSON/Tests/Benchmark/JsonLoaderBench.php

@@ -22,7 +22,7 @@ public function __construct()
     {
         $this->context = flow_context(config());
         $this->outputPath = \tempnam(\sys_get_temp_dir(), 'etl_json_loader_bench') . '.json';
-        $this->rows = (new FakeStaticOrdersExtractor(10_000))->toRows();
+        $this->rows = (new FakeStaticOrdersExtractor(1_000))->toRows();
     }
 
     public function __destruct()
@@ -34,7 +34,7 @@ public function __destruct()
         \unlink($this->outputPath);
     }
 
-    public function bench_load_10k() : void
+    public function bench_load_1k() : void
     {
         to_json($this->outputPath)->load($this->rows, $this->context);
     }

src/adapter/etl-adapter-parquet/tests/Flow/ETL/Adapter/Parquet/Tests/Benchmark/ParquetLoaderBench.php

@@ -22,7 +22,7 @@ public function __construct()
     {
         $this->context = flow_context(config());
         $this->outputPath = \tempnam(\sys_get_temp_dir(), 'etl_parquet_loader_bench') . '.parquet';
-        $this->rows = (new FakeStaticOrdersExtractor(10_000))->toRows();
+        $this->rows = (new FakeStaticOrdersExtractor(1_000))->toRows();
     }
 
     public function __destruct()
@@ -34,7 +34,7 @@ public function __destruct()
         \unlink($this->outputPath);
     }
 
-    public function bench_load_10k() : void
+    public function bench_load_1k() : void
     {
         to_parquet($this->outputPath)->load($this->rows, $this->context);
     }

src/adapter/etl-adapter-postgresql/tests/Flow/ETL/Adapter/PostgreSql/Tests/Benchmark/PostgreSqlExtractorBench.php

@@ -14,17 +14,13 @@
 #[Groups(['extractor'])]
 final class PostgreSqlExtractorBench
 {
-    private const TABLE_NAME = 'benchmark_orders_extractor';
+    private const string TABLE_NAME = 'benchmark_orders_extractor';
 
     private Client $client;
 
     public function __construct()
     {
-        $dsn = \getenv('PGSQL_DATABASE_URL');
-
-        if (!$dsn) {
-            throw new \RuntimeException('PGSQL_DATABASE_URL environment variable is not set');
-        }
+        $dsn = \getenv('PGSQL_DATABASE_URL') ?: throw new \RuntimeException('PGSQL_DATABASE_URL environment variable is not set');
 
         if (!\extension_loaded('pgsql')) {
             throw new \RuntimeException('ext-pgsql is not available');
@@ -50,37 +46,37 @@ public function __destruct()
         $this->client->close();
     }
 
-    public function bench_extract_10k_cursor() : void
+    public function bench_extract_1k_cursor() : void
     {
         $context = flow_context(config());
 
         foreach (from_pgsql_cursor(
             $this->client,
             select(star())->from(table(self::TABLE_NAME)),
-        )->withFetchSize(1000)->extract($context) as $rows) {
+        )->withFetchSize(100)->extract($context) as $rows) {
         }
     }
 
-    public function bench_extract_10k_keyset() : void
+    public function bench_extract_1k_keyset() : void
     {
         $context = flow_context(config());
 
         foreach (from_pgsql_key_set(
             $this->client,
             select(star())->from(table(self::TABLE_NAME)),
             pgsql_pagination_key_set(pgsql_pagination_key_asc('index')),
-        )->withPageSize(1000)->extract($context) as $rows) {
+        )->withPageSize(100)->extract($context) as $rows) {
         }
     }
 
-    public function bench_extract_10k_limit_offset() : void
+    public function bench_extract_1k_limit_offset() : void
     {
         $context = flow_context(config());
 
         foreach (from_pgsql_limit_offset(
             $this->client,
             select(star())->from(table(self::TABLE_NAME))->orderBy(asc(col('index'))),
-        )->withPageSize(1000)->extract($context) as $rows) {
+        )->withPageSize(100)->extract($context) as $rows) {
         }
     }
 
@@ -104,11 +100,8 @@ private function setupDatabase() : void
                 ->column(column('items', data_type_jsonb()))
         );
 
-        $extractor = new FakeStaticOrdersExtractor(10_000);
-        $context = flow_context(config());
-
         df()
-            ->read($extractor)
+            ->read(new FakeStaticOrdersExtractor(1_000))
             ->write(to_pgsql_table($this->client, self::TABLE_NAME))
             ->run();
     }

src/adapter/etl-adapter-postgresql/tests/Flow/ETL/Adapter/PostgreSql/Tests/Benchmark/PostgreSqlLoaderBench.php

@@ -15,7 +15,7 @@
 #[Groups(['loader'])]
 final class PostgreSqlLoaderBench
 {
-    private const TABLE_NAME = 'benchmark_orders_loader';
+    private const string TABLE_NAME = 'benchmark_orders_loader';
 
     private Client $client;
 
@@ -25,11 +25,7 @@ final class PostgreSqlLoaderBench
 
     public function __construct()
     {
-        $dsn = \getenv('PGSQL_DATABASE_URL');
-
-        if (!$dsn) {
-            throw new \RuntimeException('PGSQL_DATABASE_URL environment variable is not set');
-        }
+        $dsn = \getenv('PGSQL_DATABASE_URL') ?: throw new \RuntimeException('PGSQL_DATABASE_URL environment variable is not set');
 
         if (!\extension_loaded('pgsql')) {
             throw new \RuntimeException('ext-pgsql is not available');
@@ -43,7 +39,7 @@ public function __construct()
             pgsql_connection_dsn($dsn),
             mapper: pgsql_mapper(),
         );
-        $this->rows = (new FakeStaticOrdersExtractor(10_000))->toRows();
+        $this->rows = (new FakeStaticOrdersExtractor(1_000))->toRows();
         $this->context = flow_context();
     }
 
@@ -77,9 +73,9 @@ public function setUp() : void
     }
 
     #[BeforeMethods('setUp')]
-    public function bench_load_10k() : void
+    public function bench_load_1k() : void
     {
-        foreach ($this->rows->chunks(1_000) as $chunk) {
+        foreach ($this->rows->chunks(100) as $chunk) {
             to_pgsql_table($this->client, self::TABLE_NAME)->load($chunk, $this->context);
         }
     }

src/adapter/etl-adapter-text/tests/Flow/ETL/Adapter/Text/Tests/Benchmark/TextLoaderBench.php

@@ -5,7 +5,7 @@
 namespace Flow\ETL\Adapter\Text\Tests\Benchmark;
 
 use function Flow\ETL\Adapter\Text\to_text;
-use function Flow\ETL\DSL\{config, flow_context};
+use function Flow\ETL\DSL\{config, flow_context, row};
 use Flow\ETL\{FlowContext, Row, Rows, Tests\Double\FakeStaticOrdersExtractor};
 use PhpBench\Attributes\Groups;
 
@@ -22,8 +22,8 @@ public function __construct()
     {
         $this->context = flow_context(config());
         $this->outputPath = \tempnam(\sys_get_temp_dir(), 'etl_txt_loader_bench') . '.txt';
-        $this->rows = (new FakeStaticOrdersExtractor(10_000))->toRows()->map(
-            fn (Row $r) => \Flow\ETL\DSL\row($r->get('order_id'))
+        $this->rows = (new FakeStaticOrdersExtractor(1_000))->toRows()->map(
+            fn (Row $r): Row => row($r->get('order_id'))
         );
     }
 
@@ -36,7 +36,7 @@ public function __destruct()
         \unlink($this->outputPath);
     }
 
-    public function bench_load_10k() : void
+    public function bench_load_1k() : void
     {
         to_text($this->outputPath)->load($this->rows, $this->context);
     }

src/core/etl/tests/Flow/ETL/Tests/Benchmark/EntryFactory/EntryFactoryBench.php

@@ -50,10 +50,6 @@ public function provideRows() : \Generator
             'notes' => "Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s, when an unknown printer took a galley of type and scrambled it to make a type specimen book. It has survived not only five centuries, but also the leap into electronic typesetting, remaining essentially unchanged. It was popularised in the 1960s with the release of Letraset sheets containing Lorem Ipsum passages, and more recently with desktop publishing software like Aldus PageMaker including versions of Lorem Ipsum.",
         ];
 
-        yield '10k' => ['rows' => \array_map($callback, \range(1, 10_000))];
-
-        yield '5k' => ['rows' => \array_map($callback, \range(1, 5000))];
-
         yield '1k' => ['rows' => \array_map($callback, \range(1, 1000))];
     }
 }

src/core/etl/tests/Flow/ETL/Tests/Benchmark/RowsBench.php

@@ -12,26 +12,16 @@
 #[Groups(['building_blocks'])]
 final class RowsBench
 {
-    private Rows $rows;
-
     private Rows $rows100;
 
     private Rows $rows1k;
 
     public function setUp() : void
     {
-        $this->rows = (new FakeStaticOrdersExtractor(10_000))->toRows();
         $this->rows1k = (new FakeStaticOrdersExtractor(1_000))->toRows();
         $this->rows100 = (new FakeStaticOrdersExtractor(100))->toRows();
     }
 
-    public function bench_chunk_1_000_on_10k() : void
-    {
-        foreach ($this->rows->chunks(1_000) as $chunk) {
-
-        }
-    }
-
     public function bench_diff_left_100_on_1k() : void
     {
         $this->rows1k->diffLeft($this->rows100);
@@ -91,7 +81,7 @@ public function bench_partition_by_on_1k() : void
 
     public function bench_schema_on_1k_identical_rows() : void
     {
-        $this->rows->schema();
+        $this->rows1k->schema();
     }
 
     public function bench_sort_asc_on_1k() : void

src/core/etl/tests/Flow/ETL/Tests/Benchmark/Transformer/RenameEachEntryTransformerBench.php

@@ -31,7 +31,7 @@ public function setUp() : void
         $this->context = flow_context(config());
     }
 
-    public function bench_transform_10k_rows() : void
+    public function bench_transform_1k_rows() : void
     {
         (new RenameEachEntryTransformer(rename_style(StringStyles::KEBAB)))->transform($this->rows, $this->context);
     }

src/core/etl/tests/Flow/ETL/Tests/Benchmark/Transformer/RenameEntryTransformerBench.php

@@ -26,13 +26,13 @@ public function setUp() : void
                 ['id' => 3, 'random' => false, 'text' => null, 'from' => 666],
                 ['id' => 4, 'random' => true, 'text' => null, 'from' => 666],
                 ['id' => 5, 'random' => false, 'text' => null, 'from' => 666],
-            ], \range(0, 10_000))),
+            ], \range(0, 1_000))),
             flow_context(config())->entryFactory(),
         );
         $this->context = flow_context(config());
     }
 
-    public function bench_transform_10k_rows() : void
+    public function bench_transform_1k_rows() : void
     {
         (new RenameEntryTransformer('from', 'to'))->transform($this->rows, $this->context);
     }