Ok I changed my connection string to use the IP of the mySql box and added my workstation to the hosts file on the server. Same problem. The tests ran at the same slow pace.
I am really curious, what was the logic of adding the workstation ip to the hosts file on the server?
Larry Lowry
----- Original Message ----- From: "Donny Simonton" <[EMAIL PROTECTED]>
To: "'Larry Lowry'" <[EMAIL PROTECTED]>; "'Peter J Milanese'" <[EMAIL PROTECTED]>; <mysql@lists.mysql.com>
Sent: Saturday, January 22, 2005 7:34 PM
Subject: RE: Connection performance, suggestions?
Attempt to connect to mysql via ip address, and make sure on the mysql box that that you add the connecting boxes to your hosts file. Your problem should go away then.
Donny
-----Original Message----- From: Larry Lowry [mailto:[EMAIL PROTECTED] Sent: Saturday, January 22, 2005 9:25 PM To: Peter J Milanese; mysql@lists.mysql.com Subject: Re: Connection performance, suggestions?
Actually I was just trying to see why the same code to mySql is so much slower than SQL Server. I tried the same code using the data provider from CoreLabs and found the connections occur 4 times faster than the mySQL AB connector. The slowness just seems to be in the connector.
Now I know my method seems silly but to my boss it makes sense. We have web sites that make these connections. They are obviously stateless and each makes it's own connection. Where it takes 22 seconds to connect to mySql 100 times I can in the same loop execute a select statement and bring back a 700k binary field adding only abour .5 seconds to the whole loop time. Half a second to dig into a database with 400k records and yank the record 100 times but 22seconds just to open and close the conection. Something just does not seem right.
Thanks for your input.
Larry Lowry
----- Original Message ----- From: "Peter J Milanese" <[EMAIL PROTECTED]>
To: "Larry Lowry" <[EMAIL PROTECTED]>
Sent: Saturday, January 22, 2005 10:22 AM
Subject: Re: Connection performance, suggestions?
You are not simulating. Your scripting produces 100 linear connections. You
are attempting, by theory, to simulate simultaneous connections. Big
difference.
----------------- Sent from my NYPL BlackBerry Handheld.
----- Original Message ----- From: "Larry Lowry" [EMAIL PROTECTED] Sent: 01/22/2005 10:36 AM To: "Peter J Milanese" <[EMAIL PROTECTED]> Subject: Re: Connection performance, suggestions?
----- Original Message ----- From: "Peter J Milanese" <[EMAIL PROTECTED]>
To: "Larry Lowry" <[EMAIL PROTECTED]>; "mysql" <mysql@lists.mysql.com>
Sent: Friday, January 21, 2005 12:57 PM
Subject: Re: Connection performance, suggestions?
A single transaction logs into the db several times?Not really. I am simulating the connection process of a bunch of browser based/web services transactions.Correct.
I assume its a browser based transaction, no?
Are you limiting connections (in my.conf)? Have you tuned the config, if yes how so?
Here is my My.INI.
[mysqld] basedir=C:/MySql datadir=D:/MySqlData language=C:/MySql/share/english port=3306 key_buffer_size=512M table_cache=64 net_buffer_length=1M max_allowed_packet=3M query_cache_limit = 1M query_cache_size = 32M query_cache_type = 1 read_buffer_size=2M read_rnd_buffer_size=8M skip-innodb
Larry
------Original Message------ From: Larry Lowry To: mysql Sent: Jan 21, 2005 12:49 PM Subject: Connection performance, suggestions?
Before I ask this question I must state I love mySql and want to use it
more.
But I work in a shop were we mostly use SQL Server 2000. In trying to use
mySql (4.12 and 4.19) we are seeing some performance issues. I hope
someone can help me with this. I am in a Windows 2003 server environment
and
running mostly asp.net applications.
I have narrowed the problem down to the speed at which the database
connections
open and close. The following code opens a mySql database 100 times.
This takes 21 to 23 seconds. I know I should only open the connection
once
but this represents the asp environment where we are using a cluster of
web
servers. MySQL is running on an Athlon 64 3500+ with 2 gigs of memory. It
is the
only process on the server.
The SQL server code below does the same thing except to Sql Server 2000. This only takes .1 to .4 seconds. Sql Server is running on a PIII at 1.2ghz with 1gb memory. It's a crappy old box for testing.
I am using the stock Framework 1.1 SqlClient and MySql.Data.MySqlClient version 1.0.3.31712. All test boxes are on the same network switch.
Can anyone help explain this difference? Is it the database engines or the data connectors? Any help would be appreciated.
Thanks
Larry Lowry
'MySql Code Dim sDBCS As String = "Server=DB;UserId=userid;Password=pass;Database=images" Dim i As Long Dim ti As Long = Microsoft.VisualBasic.Timer()
Dim db As MySqlConnection db = New MySqlConnection(sDBCS) For i = 1 To 100 db.Open() db.Close() Next db.Dispose() tbEnd.Text = Microsoft.VisualBasic.Timer() - ti
'SQL Server
Dim sDBCS As String = "Data Source=db;User Id=userid;Password=pass;Initial
Catalog=images"
Dim i As Long
Dim ti As Long = Microsoft.VisualBasic.Timer()
Dim db As SqlConnection db = New SqlConnection(sDBCS) For i = 1 To 100 db.Open() db.Close() Next db.Dispose() tbEnd.Text = Microsoft.VisualBasic.Timer() - ti
----------------- Sent from my NYPL BlackBerry Handheld.
-- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
-- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
-- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
