db-derby-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From spykee <toma.georg...@yahoo.com>
Subject Re: Derby replication system - Need help
Date Sun, 23 Mar 2014 21:22:59 GMT
Hi,

>From my application tables ( for each table I have a trigger for INSERT,
UPDATE and DELETE operation), I will need two important information:
a) the columns names which changed when an update SQL occured.
b) the columns values that changed during an UPDATE SQL script.
 The points a) + b) will be used to create a message to be sent on the
QUEUE.

But with Derby I encountered few "issues" .

1. Is not possible to fetch only the updated columns from an UPDATE trigger.
2. There is no  way of fetching the columns name from a trigger(That's, I
want to know from an UPDATE trigger the updated columns names and to use
this information on building the message for my replication system)

I thought I can avoid using a lot of Java code for these tasks.
Using Java code for searching(filtering) only the updated columns names from
a specific table, and their values will cause me a delay on my replication
system, a drawback.

First a trigger will fire, then a Java code ( stored procedure ) will be
called ( the Java code will go back and query the db - I don't like this,
why I can't solve  this problem using Derby functionality, triggers ? ), a
comparison will be made on the last 2 rows from the audit table, pick only
the different values + the columns names, create a specific message with
this information.

There is no other ways of avoiding using Java code for these tasks ? 

Please advice me.

Regards,
George




--
View this message in context: http://apache-database.10148.n7.nabble.com/Re-Derby-replication-system-Need-help-tp138003p138203.html
Sent from the Apache Derby Users mailing list archive at Nabble.com.

Mime
View raw message