The general rule of thumb is that connection creation/destruction is the most costly 
part of approach #2.  However, it is unlikely that each connection needs a dedicated 
connection, and therefore memory resources on the server for the entirity of the 
client's lifespan... 

if your clients are firing off many queries back-to-back, buy a lot of memory for your 
server, and go w/ your option #1.

if it's a typical app, where the user queries data, looks through it for a bit, then 
queries more, yadda, yadda, yadda, then what you're probably going to want to do is 
#2.  

Without a central 'app server' as such, you can't really take advantage of connection 
pooling.  The closest you can do is write into your app that if the connection is idle 
for x amount of time, then it closes the connection, and reestablishes it upon next 
data request.

Hope this helps...

Dan Greene



> -----Original Message-----
> From: Director General: NEFACOMP [mailto:[EMAIL PROTECTED]
> Sent: Thursday, September 25, 2003 11:57 AM
> To: [EMAIL PROTECTED]; [EMAIL PROTECTED]
> Subject: Ideas on creating connections
> 
> 
> Hi groups,
> 
> I need some experienced users to guide me on this issue.
> I am developping an Application that will be accessing a 
> MySQL database through MyODBC.
> The application is developped using Visual Basic.
> 
> Since my application will be used simultaneously by more than 
> 200 users, I want to know the implication of using one of the 
> following methods for connecting to the server:
> 
> 1. At logon, I create a connection to the server and maintain 
> it throught the application life. By the application life I 
> mean that the connection stays open as long as the 
> application is loaded in the memory. So, for a user who is 
> online for 3 hours, the connection is on for those three 
> hours, even if the user goes out for a coffee.
> 
> 2. I create a connection only when a query is about to be 
> sent to the server. That means, I create a connection when I 
> want to create an ADO recordset and run a query against the 
> server. After running my query, I distroy the connection. 
> 
> 
> Currently I am using the first option since I don't exactly 
> what it is required (time and resources) to make a connection 
> to the server. Will you please tell me how fast is to connect 
> to the server. If you advise me that this method is the best, 
> I will add functions to reconnect a dead connection.
> 
> 
> Any ideas and advices are highly welcomed
> 
> 
> Thanks,
> __________________________________
> NZEYIMANA Emery Fabrice
> NEFA Computing Services, Inc.
> P.O. Box 5078 Kigali
> Office Phone: +250-51 11 06
> Office Fax: +250-50 15 19
> Mobile: +250-08517768
> Email: [EMAIL PROTECTED]
> http://www.nefacomp.net/
> 

--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:    http://lists.mysql.com/[EMAIL PROTECTED]

Reply via email to