I've been planning on making a change on the trunk code to optimize this
by adding a 'read-ahead buffer' to JAWS. This is a bit non-trivial,
since the buffer needs to be per-transaction, and some thought also
needs to be down on how this effects transaction isolation levels.
-danch
Sacha Labourey wrote:
> Thanks Danch!
>
> (even if the answer does not make me happy ;) )
>
>
> Cheers,
>
>
>
> Sacha
>
>
>
>> -----Message d'origine-----
>> Finders currently hit the database n+1 times (where N is the number of
>> beans selected by the query).
>>
>> -danch
>>
>> Sacha Labourey wrote:
>>
>>> Hello,
>>>
>>> I have question regarding the way finders work in the current
>>> implementation.
>>>
>>> When using CMP and invoking a ejbFindXXX method returning a
>>
>> collection of
>>
>>> beans, if I then invoke a method on each of these beans, will they each
>>> require a DB call (select * from xxx where PK = x) or does the
>>
>> ejbFindXXX
>>
>>> call cache the DB information that can be reused for these
>>
>> subsequents bean
>>
>>> calls (during the whole transaction)?
>>
>
>
> _______________________________________________
> JBoss-user mailing list
> [EMAIL PROTECTED]
> http://lists.sourceforge.net/lists/listinfo/jboss-user
_______________________________________________
JBoss-user mailing list
[EMAIL PROTECTED]
http://lists.sourceforge.net/lists/listinfo/jboss-user
