|
Performance
Here are some sample queries to give an idea of the performance. All queries were ran connecting to a local mysql server. Updated: July 27 2008. Tests ran against revision 130 I ran all these queries on a MacBook Pro, 2.4GHz Core 2, 4GB Ram, 7200rpm Drive. All queries ran against the following table selecting all columns. CREATE TABLE `employees` ( `employeeID` int(11) NOT NULL auto_increment, `username` varchar(255) character set latin1 default NULL, `enabled` tinyint(1) default NULL, `hourlyWage` decimal(8,2) default NULL, `startDate` date default NULL, `shiftStartTime` time default NULL, `photo` longblob, `createDate` datetime default NULL, `modifyDate` timestamp NOT NULL default '0000-00-00 00:00:00' on update CURRENT_TIMESTAMP, PRIMARY KEY (`employeeID`) ) ENGINE=InnoDB DEFAULT CHARSET=utf8; [DEBUG] com.maclema.mysql::QueryHandler : Mysql Result [DEBUG] com.maclema.mysql::QueryHandler : Rows: 10 [DEBUG] com.maclema.mysql::QueryHandler : Query Size: 1.48 KB [DEBUG] com.maclema.mysql::QueryHandler : Total TX: 1.57 KB [DEBUG] com.maclema.mysql::QueryHandler : Query Time: 19 ms [DEBUG] com.maclema.mysql::QueryHandler : Mysql Result [DEBUG] com.maclema.mysql::QueryHandler : Rows: 100 [DEBUG] com.maclema.mysql::QueryHandler : Query Size: 9.11 KB [DEBUG] com.maclema.mysql::QueryHandler : Total TX: 10.68 KB [DEBUG] com.maclema.mysql::QueryHandler : Query Time: 16 ms [DEBUG] com.maclema.mysql::QueryHandler : Mysql Result [DEBUG] com.maclema.mysql::QueryHandler : Rows: 1000 [DEBUG] com.maclema.mysql::QueryHandler : Query Size: 87.06 KB [DEBUG] com.maclema.mysql::QueryHandler : Total TX: 97.75 KB [DEBUG] com.maclema.mysql::QueryHandler : Query Time: 87 ms [DEBUG] com.maclema.mysql::QueryHandler : Mysql Result [DEBUG] com.maclema.mysql::QueryHandler : Rows: 10000 [DEBUG] com.maclema.mysql::QueryHandler : Query Size: 878.22 KB [DEBUG] com.maclema.mysql::QueryHandler : Total TX: 975.98 KB [DEBUG] com.maclema.mysql::QueryHandler : Query Time: 346 ms [DEBUG] com.maclema.mysql::QueryHandler : Mysql Result [DEBUG] com.maclema.mysql::QueryHandler : Rows: 100000 [DEBUG] com.maclema.mysql::QueryHandler : Query Size: 8,878.23 KB [DEBUG] com.maclema.mysql::QueryHandler : Total TX: 9,854.22 KB [DEBUG] com.maclema.mysql::QueryHandler : Query Time: 3526 ms [DEBUG] com.maclema.mysql::QueryHandler : Mysql Result [DEBUG] com.maclema.mysql::QueryHandler : Rows: 200000 [DEBUG] com.maclema.mysql::QueryHandler : Query Size: 17,864.99 KB [DEBUG] com.maclema.mysql::QueryHandler : Total TX: 27,719.21 KB [DEBUG] com.maclema.mysql::QueryHandler : Query Time: 7340 ms |
► Sign in to add a comment
would you please post the performance of these same queries ran directly from a mySQL client? There is nothing to compare this info to, so it is hard to get an idea of performance similarities or losses. Thanks.
I concur with instantkevin. I have been scowering the Internet for a direct comparison and have nothing to show for it. It's great that you went to the effort of performing and publishing these figures, but without a comparison they are of limited use.
Also, can you publish the code you used for the bench-marking? Thanks.
okay...... i can compare it to AMFPHP, and i dunno whats wrong with my AMFPHP but it struggles hardly to transmit more than 1000 entries.
here are my results: (testing HTTPService and AMFPHP)
SPEED COMPARISON
aktuell gibt es folgende methoden:
Post & Get:
1. XML over HTTP
CLIENT SPEED: 120
SERVER SPEED: 120
(from 0 to 300)
DATA COMPRESSION: none
RELIABLE: 3
(0 = not reliable, 5 = very reliable)
2. AMFPHP over HTTP
CLIENT SPEED: 180
SERVER SPEED: 50
(from 0 to 300)
DATA COMPRESSION: yes
RELIABLE: 3
(0 = not reliable, 5 = very reliable)
READING 1670 ENTRIES FROM A MYSQL TABLE VIA
php script that returns a XML file via POST & GET: 16: START WORK 3.7.2009 20:40:30 17: STOP WORK 3.7.2009 20:40:36 6sec.
AMFPHP: 26sec.
i would like to see some blazeDS running on different java-server-installations benchmark results!
"I am wondering if anyone has an idea about increasing blazeds performance. I am receiving information from a database for a chart. Each record is a java object and the entire set of aproximately 180K records is contained in two separate arrays since it is a multi-series chart. I convert the records into actionscript objects from java objects when I get them. >From the time the records are sent to the flex app, til the time they get >drawn on the screen takes approximately 45+ seconds. This seems pretty long to me. Is this on par with what should be expected? If not, what could I do to improve this performance?"
I thought AMFPHP is supposed to be faster because it transmits in binary... hmmm.