PHP MySQLi mysqli_poll() - Asynchronous Connection Management

mysqli_poll()

The mysqli_poll() function is used to examine (poll) a list of connections on which we initiated asynchronous queries using the MYSQLI_ASYNC flag. In the code, we will open 10 parallel connections that will read this data simultaneously, and the mysqli_poll() function will check which one finished first.

EXAMPLE

<?php
$sql = "SELECT * FROM mysqli_table";
$all_connections = array();

// Open 10 connections and initiate asynchronous queries
for ($i = 0; $i < 10; $i++) {
    $conn = mysqli_connect("localhost", "root", "", "mysqli_database");
    // MYSQLI_ASYNC tells the database to immediately return control to PHP while searching for data in the background
    $conn->query($sql, MYSQLI_ASYNC);
    $all_connections[] = $conn;
}

$processed = 0;
do {
    $connections = $errors = $reject = $all_connections;
    
    // Polling connections (timeout set to 1 second)
    // NOTE: The mysqli_poll() function checks which of the asynchronous connections have finished 
    // fetching data and are ready to return results. The $connections, $errors, and $reject variables 
    // are automatically populated with the corresponding objects.
    if (!mysqli_poll($connections, $errors, $reject, 1)) {
        continue;
    }

    foreach ($connections as $conn) {
        // reap_async_query() retrieves the prepared result from an asynchronously completed query
        if ($result = $conn->reap_async_query()) {
            // Output the raw first row using print_r
            print_r($result->fetch_row()); echo "<br>";
            echo "<br>";
            mysqli_free_result($result);
        } else {
            die(mysqli_error($conn));
        }
        $processed++;
    }
} while ($processed < count($all_connections));

foreach ($all_connections as $conn) {
    mysqli_close($conn);
}
?>

RESULT

The result will be displayed here...

Practical use: This is an ultimate optimization technique. Instead of PHP waiting for the response of one query before sending the next, it sends them all at once. This drastically shortens page load times when working with large amounts of data.

Parallel execution of 10 or more queries simultaneously.

Using the reap_async_query() method to retrieve results.

Significantly increasing application responsiveness during complex operations.