Ashley Sheridan wrote:
> On Wed, 2008-10-29 at 19:57 -0500, Micah Gersten wrote:
>   
>> Ashley Sheridan wrote:
>>     
>>> On Wed, 2008-10-29 at 19:49 -0500, Micah Gersten wrote:
>>>   
>>>       
>>>> Ashley Sheridan wrote:
>>>>     
>>>>         
>>>>> On Wed, 2008-10-29 at 19:43 -0500, Micah Gersten wrote:
>>>>>   
>>>>>       
>>>>>           
>>>>>> Ashley Sheridan wrote:
>>>>>>     
>>>>>>         
>>>>>>             
>>>>>>> On Wed, 2008-10-29 at 19:25 -0500, Micah Gersten wrote:
>>>>>>>   
>>>>>>>       
>>>>>>>           
>>>>>>>               
>>>>>>>> Ashley Sheridan wrote:
>>>>>>>>     
>>>>>>>>         
>>>>>>>>             
>>>>>>>>                 
>>>>>>>>> On Thu, 2008-10-30 at 08:55 +1100, Chris wrote:
>>>>>>>>>   
>>>>>>>>>       
>>>>>>>>>           
>>>>>>>>>               
>>>>>>>>>                   
>>>>>>>>>> Waynn Lue wrote:
>>>>>>>>>>     
>>>>>>>>>>         
>>>>>>>>>>             
>>>>>>>>>>                 
>>>>>>>>>>                     
>>>>>>>>>>> I sent an email to the mysql list, but it reminded me of a question 
>>>>>>>>>>> I had
>>>>>>>>>>> for people structuring their PHP code.  What's the general way that 
>>>>>>>>>>> people
>>>>>>>>>>> structure their connections?  Right now, I spawn off two 
>>>>>>>>>>> mysql_connect calls
>>>>>>>>>>> at the top of the file that includes my database calls, using 
>>>>>>>>>>> "true" for the
>>>>>>>>>>> fourth parameters, so as to create two new connections.  Then I use 
>>>>>>>>>>> those
>>>>>>>>>>> two connections for two different databases I have to query from.
>>>>>>>>>>>
>>>>>>>>>>> Is it better just to use mysql_select_db within the query function 
>>>>>>>>>>> every
>>>>>>>>>>> time for the same connection?  Should I use mysql_connect every 
>>>>>>>>>>> time without
>>>>>>>>>>> using "true", so as to re-use connections.  Should I be using 
>>>>>>>>>>> pconnect
>>>>>>>>>>> instead?
>>>>>>>>>>>
>>>>>>>>>>> I spent some time looking for answers to these questions, but am 
>>>>>>>>>>> getting
>>>>>>>>>>> conflicting answers.  Some people think relying on the re-use of 
>>>>>>>>>>> these
>>>>>>>>>>> functions is good, some think that explicit management is better.  
>>>>>>>>>>> In
>>>>>>>>>>> general, how have people on the list found them?  For example, is 
>>>>>>>>>>> having
>>>>>>>>>>> constant mysql_select_db calls a problem?
>>>>>>>>>>>           
>>>>>>>>>>>               
>>>>>>>>>>>                   
>>>>>>>>>>>                       
>>>>>>>>>> Are they connecting as the same user and on the same server? Then 
>>>>>>>>>> you 
>>>>>>>>>> can replace with a mysql_select_db call.
>>>>>>>>>>
>>>>>>>>>> If they aren't both of those, you have no choice.
>>>>>>>>>>
>>>>>>>>>> No idea if it'll make much of a difference (performance wise etc) 
>>>>>>>>>> but 
>>>>>>>>>> I'd leave it as two connections.
>>>>>>>>>>
>>>>>>>>>>         
>>>>>>>>>>             
>>>>>>>>>>                 
>>>>>>>>>>                     
>>>>>>>>> How difficult would it be to converge the 2 databases into one? This
>>>>>>>>> would obviously use less memory (not sure exactly how big the 
>>>>>>>>> footprint
>>>>>>>>> of each connection is though) and will slightly speed up page display
>>>>>>>>> time (as you only have to wait for one connection to be made rather 
>>>>>>>>> than
>>>>>>>>> two)
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Ash
>>>>>>>>> www.ashleysheridan.co.uk
>>>>>>>>>
>>>>>>>>>   
>>>>>>>>>       
>>>>>>>>>           
>>>>>>>>>               
>>>>>>>>>                   
>>>>>>>> Generally you want separation of data.  MySQL doesn't have a problem
>>>>>>>> accessing another DB on the same server with the same connection.  
>>>>>>>> Also,
>>>>>>>> how would database convergence use less memory?
>>>>>>>>
>>>>>>>> .
>>>>>>>> Thank you,
>>>>>>>>     
>>>>>>>>         
>>>>>>>>             
>>>>>>>>                 
>>>>>>> For arguments sake, open 1000 database connections, all to different
>>>>>>> databases. Now tell me that each connection doesn't have a footprint. At
>>>>>>> the end of the day, whist it may seem fine for a script to have 2
>>>>>>> connections open, the least open the better. Imagine 100 users
>>>>>>> simultaneously accessing a page that opens 10 connections. Suddenly you
>>>>>>> have 200 connections open, not a great idea. If you could amalgamate the
>>>>>>> db's, you'd have half as many connections open.
>>>>>>>
>>>>>>> If you're still having trouble understanding why having two database
>>>>>>> connections open is bad (regardless of whether they are on the same
>>>>>>> server or not) the I think web development is the wrong career for you.
>>>>>>>
>>>>>>>
>>>>>>> Ash
>>>>>>> www.ashleysheridan.co.uk
>>>>>>>   
>>>>>>>       
>>>>>>>           
>>>>>>>               
>>>>>> The answer in your case is not to combine the DBs necessarily, but
>>>>>> consolidate the connections used.  Like I said, you can use 2 MySQL DBs
>>>>>> on the same connection in PHP.  There's no reason to sacrifice
>>>>>> separation of data.
>>>>>>
>>>>>> Thank you,
>>>>>> Micah Gersten
>>>>>>     
>>>>>>         
>>>>>>             
>>>>> I'm sure if you look at the OP codes on your suggestion, you'd still use
>>>>> the same memory as having two separate connections open, unless you
>>>>> closed one first. Thing is, opening and closing database connections has
>>>>> its own overheads.
>>>>>
>>>>>
>>>>> Ash
>>>>> www.ashleysheridan.co.uk
>>>>>   
>>>>>       
>>>>>           
>>>> How is using one connection the same as having 2 open?  You just change
>>>> databases if you want to, or use the fully qualified table name
>>>> (database.table) in your query.  What extra overhead is there in that?
>>>>
>>>> Thank you,
>>>> Micah Gersten
>>>>         
>>> Having one connection open at a time has only the overhead of the
>>> opening and closing of connections. As far as I know, you can't have two
>>> databases open on one connection, but please correct me if I am wrong.
>>>
>>>
>>> Ash
>>> www.ashleysheridan.co.uk
>>>   
>>>       
>> With MySQL, you can change the DB from query to query with
>> mysql_select_db.  The alternative as I stated in my last post is to use
>> the fully qualified table name (database.table) in your query.  MySQL
>> doesn't care which DB you have open if you do that.  In both of these
>> cases, the same connection is used.
>>
>> Thank you,
>> Micah Gersten
>>     
> And you can say for certain that there's no extra overhead involved?
>
>
> Ash
> www.ashleysheridan.co.uk
>   

With the fully qualified table name I can say for sure.  I can only hope
with the mysql_select_db that it just issues a use dbName command.

Thank you,
Micah Gersten
onShore Networks
Internal Developer
http://www.onshore.com




-- 
PHP General Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php

Reply via email to