开发者

"caching" an array of database values in php (2 options, which is better?)

开发者 https://www.devze.com 2023-01-09 07:30 出处:网络
(I use the word var in the following because I\'m not sure what it is I think it\'s a global variable for the class)

(I use the word var in the following because I'm not sure what it is I think it's a global variable for the class)

I want to cache values that are gotten from a table in a database, to save repeated queries. What I've done is create a singleton class which has a function to check a var within the class if a key exist, if it does then just return the value from the var, if it doesn't then query the database and load it into the var and return it.

Or would a better way of doing it be using a global variable (does a global variable have the same value no matter from which file accesses it?) and not using a class at all.

Or don't bother caching and just query the database every time the va开发者_JAVA百科lue is needed.

Should I post the source of what I've come up with?


Check out memcached. Its purpose is to do exactly what you just described. It takes a database query and stores the results in memory which has much faster I/O. When the same query is run you check to see if the value is stored in memory and only hit the database if its not. The value of memcached over your current solution is that its persistent across connections.


If it's just links you could store them in your session meaning they're only created the first time you browse to a page with that on it. If you are able to get all the links up front and store them in the session you can do..

session_start();
setupLinks();

// Getting a link by Link ID
if( ($aLink = getLink(562)) !== null) {
    echo "The Link Title is: {$aLink['value']}<br>";
} else {
    echo "The Link Title is unknown<br>";
}

// Getting a link by Key
if( ($aLink = getLink('my_link_key')) !== null) {
    echo "The Link Title is: {$aLink['value']}<br>";
} else {
    echo "The Link Title is unknown<br>";
}

function setupLinks() {

    // Only do this once, so if the links exist do nothing
    if(!isset($_SESSION['aLinks'])) {
        return null;
    }

    storeLinks(getAllLinks()); // your own custom function
}

function storeLinks($aLinks) {
    foreach($aLinks as $aLink) {
         $_SESSION['aLinks'][$aLink['id']] = array(
              'key'   => $aLink['key'],
              'value' => $aLink['value']
         );
    }
}

function getLink($iLinkID) {
    return isset($_SESSION['aLinks'][$iLinkID]) ? $_SESSION['aLinks'][$iLinkID] : null;
}

function getLinkByKey($sLinkKey) {
    foreach($_SESSION['aLinks'] as $aLink) {
        if($sLinkKey == $aLink['key']) { return $aLink; }
    }
    return null;
}


It very depends on your queries - how expensive and heavy are they for the database? If it's pretty simple and light queries - I would say caching is not needed (unless you would have really high amounts of them - thousands or millions in short time). For simple repetitive queries your DB probably will also use some kind of query cache - this will also speed things up.

Caching expensive queries would make more sense. Then you can use memcache as suggested, you may also look to APC cache as well.

0

精彩评论

暂无评论...
验证码 换一张
取 消

关注公众号