CPU usage scaling with file-size

[ryzr]

Given the following code:

ini_set("auto_detect_line_endings", '1');

$reader = \League\Csv\Reader::createFromPath(public_path($csv->file), 'r')
    ->setDelimiter($csv->delimeter)
    ->setEnclosure($csv->enclosure)
    ->setEscape($csv->escape);

$data = $reader->query();

foreach ($data as $row_num => $row) {
    //do nothing
}

Memory consumption is fine, achieving everything perfectly. However I see a higher CPU usage when using larger files. In a local environment:

38.8kb CSV -> Apache process reaches ~9% CPU 5.86MB CSV -> Apache process reaches ~45% CPU

Just wanting to know:

• Is this an expected outcome? I assumed everything was processed row-by-row (which seems to be the case as memory usage is fine)
• Is this something that could be rectified with proper server setup? (live setup is LNMP, local setup is WAMP)

[ryzr]

Worth noting, adding a micro sleep in the loop usleep(100); normalized CPU usage to around 10%, for many different sized CSVs. The higher the sleep interval, the lesser the CPU usage.

[nyamsprod]

Hi @ryzr sorry But I fail to reproduce this behavior on my personal laptop or on a remote server must be a configuration problem outside of the package scope