On 4/26/06 2:15 PM, "Tom Schindl" wrote:

> Joachim Zobel wrote:
>> On Tue, 18 Apr 2006 08:17:47 +0200
>> Tom Schindl wrote:
>>
>>
>>> The better approach seems to me using mod_dbd which has the draw back
>>> that it is only available on apache-2.2 at least for mysql it looks
>>> like it's not rocket science to get it working.

>>
>>
>> AFAIK this has not yet been DBIfied and the interface is not yet as
>> rich as the one provided by DBI.
>>
>> One severe limitation is at the moment that you can have only one pool
>> per apache. So having 3 db handles is not an option, at least if they
>> access the db as different users.
>>
>> Of course a db interface provided by apache is a very promising
>> approach.
>>
>> Sincerely,
>> Joachim
>>
>>

>
> Hi, I don't think you need to DBIfy this. In my thinking you only use
> the framework to pool the connection. The connection pointer fetched is
> fed into DBD::MySQL, all this is happing on the C-level and you don't
> even know that mod_dbd/apr_dbd have been involved.
>
> In my thinking you only have to teach DBD::* to use mod_dbd/apr_dbd to
> fetch the database connection instead of opening one itself.


Just to jump in here--what is the best source of DOCs for doing this? I
haven't started playing with it yet, so I have not done my homework, but was
just curious to hear what others have found to be most useful.

Thanks,
Sean