Because we want to build a very scalable application where we should be able
to buy more hardware (read more server boxes) to increase performance and
split the workload across the EJB servers. Also we will, by this approach,
be able to split the workload to different databases depending on the usage
of the data. If the gui is requesting data, the resultset is moved to the
local database. If the server is handling the data, the data stays in the
main database. The single point of truth (Oracle database) should fire small
and well defined queries. This is the main reason behind the strategy of
having more the one database.
We are still in an investigation phase, so any suggestion on doing this any
other way is very welcome!
Regards
Stefan
>Hi,
>But why two RDBMS(Sqlserver and Oracle)
>Regards
>Lakshmi Srikar
Hervén Stefan
<stefan.[Email Address Removed] To: "LazyDBA.com
Discussion"
pitalc.se> <[Email Address Removed] cc: (bcc:
lakshminathan.v/Polaris)
06/28/02 02:24 PM Subject: Distribute query
result accross
databases?
Friends,
We are having a discussion where we are trying to design an architecture to
allow distribution of query results over several databases. We have a
single
database in the middle where the "truth" is kept. When a user runs a query
it is fetched up to the gui in parts of 100 rows. When the user scrolls
down
in the result and reaches row 100, the query will fire again to retrieve
the
next 100 rows and so on. The system is built on the J2EE architecture.
The problem is that the resultset might have been changed during the scroll
(rows added or deleted). We want to keep the original resultset during the
user is looking at the data. The user needs to manually click "Refresh"
button to fire the query again. Has anoyone implemented anything like this?
We are presently discussing to :
Klient -> App Server with local SQL Server -> Database Server with Oracle
When the user queries the Oracle database for scrolling we will redirect
the
user to the SQL server on the App Server box, connect to the Oracle db
using
OLEDB in insert the complete resultset of the query in a temporary table in
the SQL Server. The user will then use only the temptable in SQL Server
when
viewing the data, until refresh is hit. Would this make sense? Anyone knows
any good articles and/or books about this?
Many thanx!
Stefan
--------
Oracle documentation is here:
http://tahiti.oracle.com/pls/tahiti/tahiti.homepage
To unsubscribe: send a blank email to oracledba-[Email Address Removed] subscribe: send a blank email to oracledba-[Email Address Removed] the list archive: http://www.LAZYDBA.com/odbareadmail.pl
Tell yer mates about http://www.farAwayJobs.com
By using this list you agree to these terms:
http://www.lazydba.com/legal.html
_____________________
Stefan Hervén
stefan.[Email Address Removed] C
Kungsbron 1, WTC
Box 70370
SE-107 24 STOCKHOLM
Tel: 08-5723 6474
Mob: 070-423 6474
> Sms: [Email Address Removed]
Oracle LazyDBA home page