Hi,
I've finally understood my mappings... :-) a bit slow but got there at the
end...
--> under AndroMDA 3.0, Hibernate 2.1.8, Spring 1.1.5
- add outer-join="true" to both in the Indivi and Dossoc mapping
files for the Indsoc relationship
- leave all the relationships to out
just open your /mda/conf/andromda.xml file (assuming you're using
an AndroMDA version later than 3.0 final) or /mda/project.xml (otherwise)
you will see a namespace for bpm4struts, just add a property called
hibernateJdbcBatchSize and give it a value between 5 and 30 (see documentation
link
wouter --> How do I do it? which file do I need to change?
_
Reply to the post : http://galaxy.andromda.org/forum/viewtopic.php?p=1564#1564
Posting to http://forum.andromda.org/ is preferred over posting to the mailing
list!
--
you can currently set that property on a per-entity basis, but you can define
it globally:
--
Wouter Zoons - [EMAIL PROTECTED]
http://www.andromda.org/
_
Reply to the post : http://galaxy.andromda.org/forum/viewtopic.php?p=1561#1561
Postin
I've added a batch size in my Indivi HBM file:
it will be nice if the batch-size is dynamic based on the query search result
counter.
How can I set the batch-size in the spring/hibernate cartridge?
I'm not sure is the best solution for this problem...
it's recommended not to use the <> stereotype, use the 'query
flag' instead
you can find this feature by right-cliking on an operation and opening the
specification
--
Wouter Zoons - [EMAIL PROTECTED]
http://www.andromda.org/
_
Reply t
I don't use the Hibernate cartridge for queries but perhaps someone else can
help:
I think you should try to get the HBM file to do what you want manually, then
it should be easy to get it to generate that way.
_
Reply to the post :
username (Identifiant) --> thales
password (Mot de passe) --> mda
_
Reply to the post : http://galaxy.andromda.org/forum/viewtopic.php?p=1537#1537
Posting to http://forum.andromda.org/ is preferred over posting to the mailing
list!
--
I haven't got a public site... can I email it to u?
And which class do you need? CriteriaSearch? CriteriaSearchConfiguration?
CriteriaSearchParameter?
_
Reply to the post : http://galaxy.andromda.org/forum/viewtopic.php?p=1530#1530
Posting
You need to put it on some website and link to it here using the [img] tags...
Can you please post the generated Criteria code?[/img]
_
Reply to the post : http://galaxy.andromda.org/forum/viewtopic.php?p=1525#1525
Posting to http://forum.an
me too I'm a bit confused... I've try to add andromda.hibernate.lazy=false at
the association end and that's produced more queries than before... :-)
Where can I send my class diagram exported as PNG? because this topic starts to
be a bit to confusing...
I don't think proxy will help because
Had you tried enabling Hibernate Proxies ?
Carlos
_
Reply to the post : http://galaxy.andromda.org/forum/viewtopic.php?p=1491#1491
Posting to http://forum.andromda.org/ is preferred over posting to the mailing
list!
I'm a bit confused with the behaviour you describe. All I can say is that the
behaviour you want is possible, but you may have to play with the mapping files
a bit.
Try to remove the outer-join="auto" attribute (or set it to false). It's not
very clear what this does for collections when laz
yes I do... so what should I do to optimise the amount of sql queries?
_
Reply to the post : http://galaxy.andromda.org/forum/viewtopic.php?p=1370#1370
Posting to http://forum.andromda.org/ is preferred over posting to the mailing
list!
--
Are you sure you are not iterating over the 'indsocs' collections after calling
Criteria.list()?
_
Reply to the post : http://galaxy.andromda.org/forum/viewtopic.php?p=1345#1345
Posting to http://forum.andromda.org/ is preferred over posting
...
...
...
As wouter pointed out, you are probably modeling 'indivi ---> indsocs' as
composition?
_
Reply to the post : http://galaxy.andromda.org/forum/viewtopic.php?p=1328#1328
Posting to http://forum.andromda.org/ is preferred over posting to the ma
My suspicion is that you have the 'indsocs' collection mapped with lazy="false"
(the default). Hibernate is then eagerly loading each element of the collection
along with some single-ended associations (acteur, portefeulle, etc).
Can you please post the mapping files for 'indivi' and 'indsocs
Luis,
08:29:33,200 INFO [SettingsFactory] Use outer join
fetching: true
First query:
08:31:24,981 INFO [STDOUT] Hibernate: select this.ID
as ID0_, this.INDI as INDI0_, this.ID_TIER as ID_TIER0_, this.ID_SITUFAMI as
ID_SITUF4_0_, this.ID_NATI as ID_NATI0_, this.ID_PORT as ID_PORT0_,
th
ced,
The solution you were given is for Hibernate 3. Can you please check in your
system logs if the use_outer_join property is set in Hibernate?
Please also post the SQL that is being generated (the select for the Criteria,
and then the first of the 100 selects).
_
It look like I need to set a "fetch mode" to subselect on my collection... how
do I do that with the hibernate cartridge? If I understood the documentation,
this behaviour "should" be enable by default with the cartridge if a finder
method returns a collection. So instead of 1 + 100 queries
ced wrote:
So if I'm doing my own query then I'm going to bypass completely the hibernate
layer and populating manually all the POJOs. So I'm correct the hibernate query
algorithm is not efficient... or have I looked in the wrong place or lost the
plot? :-)
(end of quote)
You've turned on
it depends on whether or not you told hibernate to eagerly load the
relationship (all related pojos will be created)
you can model this by setting the association to 'composition'
the default is lazy loading
--
Wouter Zoons - [EMAIL PROTECTED]
http://www.andromda.org/
_
I'm happy with the normal behaviour of the FinderMethod query but I'm a bit
confused about the relationships... Is it hibernate which do the relationship
query when an entity object is instantiated? or is it something else which do
that?
I've looked the generated code of my FinderMethod and
simple queries can easily be generated by the cartridges, but for more
complicated stuff and in order to be able to fine-tune for performance you can
override the queries, as explained here: (overriding queries)
--
Wouter Zoons - [EMAIL PROTECTED]
http://www.andromda.org/
_
I know that's what I was talking about if I was doing my own database access
layer, but the point of using AndroMDA is that the DB layer is 100% generated
by modeling entities. The thing I was asking is it a problem of algorithm
efficiency from the cartridge or from Hibernate itself?
__
why not retrieve everything in a single query ? you can join multiple tables
and have them returned in a single shot
--
Wouter Zoons - [EMAIL PROTECTED]
http://www.andromda.org/
_
Reply to the post : http://galaxy.andromda.org/forum/viewtop
I'm a newbie in term of Hibernate, so I've got a question... :-)
If I'm modeling an Entity (which has relationships with other entities) and
uses a FinderMethod with a search "like" criteria. The runtime behaviour of the
finder method is as follow:
- if the search returns 100 results
-
28 matches
Mail list logo
