Performance of fetching DB items individually vs looping with php - php

Question

Out of curiosity, which of these segments of code would have the faster performance time when implemented on a mass scale?
Let's say we have a table members and we want to fetch their photos.
Method 1
$members = $db->query('SELECT * FROM members ORDER BY id ASC');
foreach($members as $m)
$memberIDs[] = $m['id'];
$photos = $db->query('SELECT * FROM photos WHERE member IN'.join(',', $memberIDs).' ORDER BY id ASC');
foreach($photos as $p) {
// multi_arr_search(search_val, search_column, search_array) returns parent key in multi dimensional array
$memberArrayKey = multi_arr_search($p['member'], 'id', $members);
$members[$memberArrayKey]['photos'][] = $p;
}
OR
Method 2
$members = $db->query('SELECT * FROM members ORDER BY id ASC');
foreach($members as $k=>$m)
$members[$k]['photos'] = $db->query('SELECT * FROM photos WHERE member='.$m['id'].' ORDER BY id ASC');
Method 1 would result in fewer queries being ran, but requires more PHP work.

Answer 1

neither. You are using 15 years old deprecated methods.
If you want to go fast, you need PDO prepared statements.
#alberto is right. I am pretty sure, you can have one and only sql to perform this but here is a way you might want to consider in case off:
$showError = true;
define("SQLHOST", "127.0.0.1");
define("SQLUSER", "login");
define("SQLPASS", "password");
define("SQLSGBD", "database");
$conn = new PDO('mysql:host=' . SQLHOST . ';dbname=' . SQLSGBD . ';charset=UTF8', SQLUSER, SQLPASS);
$conn->setAttribute(PDO::ATTR_ERRMODE, PDO::ERRMODE_EXCEPTION);
$sql1 = 'SELECT * FROM members ORDER BY id ASC';
$stmt1 = $conn->prepare($sql1);
$sql2 = 'SELECT * FROM photos WHERE member IN ? ORDER BY id ASC';
$stmt2 = $conn->prepare($sql2);
try {
$stmt1->execute();
$members = $stmt1->fetchAll(PDO::FETCH_ASSOC);
} catch (PDOException $e) {
if ($showError === true) {
var_dump("error query 1:" . __LINE__ . "-------" . __FUNCTION__ . "-------" . $e->getMessage());
exit;
}
}
if (count($members) !== 0) {
$memberIDs = array();
foreach ($members as $m) {
$memberIDs[] = $m['id'];
}
$memberlist = join(',', $memberIDs);
foreach ($members as $memberArrayKey => $result1) {
$stmt2->bindParam(1, $memberlist, PDO::PARAM_STR);
try {
$stmt2->execute();
$photos = $stmt2->fetchAll(PDO::FETCH_ASSOC);
} catch (PDOException $e) {
if ($showError === true) {
var_dump("error query 2:" . __LINE__ . "-------" . __FUNCTION__ . "-------" . $e->getMessage());
exit;
}
}
if (count($photos) !== 0) {
$memberArrayKey = multi_arr_search($p['member'], 'id', $members);
$members[$memberArrayKey]['photos'][] = $p;
}
}
}

Answer 2

Well, hard truth: none is actually scalable. If you're working with large sets of data, PHP will need quite a bit of memory to perform all the operations to link users and photos and will do it in a much more inefficient way than the DB can.
You should have the tool you have that is best at doing these kind of things, which is your DB: the DB should join the users & photos and get you that set.. anything you do differently will mean reduced performance and scalability.

Answer 3

1- It depends on many factors how are you working with MySQL, if you connect and disconnect connection with the DB at each call.
2- MySQL Server is in the same PHP.
Ideally, connect 1 time mysql to retrieve information with a single Query. And in your case, you can solucuonar with a JOIN. Which it is best to disconnect as quickly as possible and continue with MySQL PHP.
Another recommendation is to not use " * " in the SELECT. Only puts the fields that you go to work. This eat fewer resources because only you use is requested. It is optimal in Memory, CPU, Data via LAN