What is up with Pear DB speed?

I've been using PEAR DB for quite a while now and I was a big fan until about an hour ago. I liked the object oriented programming, portability wasn't my reason for using it. I never expected that it would be significantly slower than the native mysql functions. Today I'm benchmarking some new servers that I'm setting up. The results report that using pear on on my test page is 13 times slower than the native functions for mysql. Here are my benchmark specs: I'm using http_load to test the site, the parameters of the test are to run 10 concurrent clients for 60 seconds and see how many pages are returned. The database is on a separate server so there is no CPU overhead from the database query. The result set of the query is 40 records. When I run my sample code with pear it returns an average of 24 pages a second. When I run the same query with native functions it serves 322 pages a second. Am I doing something grossly wrong with Pear DB or is it just that much slower? If this is a fact of life, how come I never seen anything mentioning that Pear DB is that much slower? There should be a big red flashing warning or something. Here are the sample pages I'm testing. <?php # With PEAR include_once 'DB.php'; include_once 'DB/mysql.php'; $db = new DB_mysql; $dsn = "mysql://user@host/db"; $db->connect (DB::parseDSN($dsn,true)); $sql = "SELECT email FROM members"; $result = $db->query($sql); while($row = $result->fetchrow(2)){ echo "$row[email]<BR>\n"; } ?> --------------------------------------------------- <?php # Native mysql_connect("host","user",""); mysql_select_db("db"); $sql = "SELECT email FROM members"; $result = mysql_query($sql); while($row = mysql_fetch_assoc($result)){ echo "$row[email]<BR>\n"; } ?> -- Jeff Bearer, RHCE Webmaster, PittsburghLIVE.com