You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@jackrabbit.apache.org by cmoulliard <cm...@gmail.com> on 2009/05/19 15:18:01 UTC

Jackrabbit as a OODBMS !!

Hi,

I would like to have the point of view of the users about the following
concern :

Can we use or consider jackrabbit as a OODBMS system to store java beans /
Pojos ?

Why : I investigate this question because in my application which is a
Straight Through Processing (STP) one, I transform external data (received
from CSV, ... files) into Java Objects which are later transformed into
external format (FIX, Swift, ...). To avoid to have too much transformation
in the STP process (IN to Internal format, internal format to text, text to
internal format, internal format to external format, ...), I think to use
the POJOs as internal format of my application and applying Java to Java
transformation to generate my internal/external format.

Obviously, I must save/store the objects somewhere to persist data in case
of application crash.

Different approaches exist :

1) Export the content of the object into String, XML, ... and save it in a
text/blob field of a RDBMS
2) Save the object in a BLOB field of a RDBMS
3) Save the objects in a OODBMS
4) idem but using JDO specification top of OODBMS, RDBMS
5) Use Java Content Repository (JCR)

The 1) is very simple to do but required additional transformations from
object to string and string to objects
The 2) seems better because the object is saved in a BLOB. What about
performance ? Is it the bast way to store objects ?
Scenario 3) 4) are interesting but they required a secondary Database (near
my MySQL DB), will complicated the transaction because I have to handle
three components in case of rollback (JMS, RDBMS and OODBMS
The 4) is perhaps mush more interesting because objects are stored in a
RDBMS and JDO is supported by Spring. I can also done queries top of my DB.
Can I use JDO to also perform my SQL queries ? Probably yes but is it
interesting ?

To use the option 5) (jackrabbit), different questions come to me like : can
I do query to retrieve my objects, can I use Spring with JCR/jackrabbit to
handle transaction and DAO layer, can I store natively my POJOs or do I have
to map them (hope this is not the case), can I use Jackrabbit in combination
with by example Mysql - RDBMS db ?

Regards

Charles Moulliard
Senior Enterprise Architect (SOA)
Apache Camel Committer

-----
Charles Moulliard
SOA Architect

My Blog :  http://cmoulliard.blogspot.com/ http://cmoulliard.blogspot.com/  
-- 
View this message in context: http://www.nabble.com/Jackrabbit-as-a-OODBMS-%21%21-tp23616237p23616237.html
Sent from the Jackrabbit - Users mailing list archive at Nabble.com.