ADB Adapter:
With ADB we can only pick up the data from one database and put it in another one or vice versa. Design time and run time environments, which are to be managed separately. Very scalable component to handle both inbound and outbound messages. ADB adapters might be useful in scenarios where we have large amount of data. ADB uses ODBC to connect, JDBC uses JDBC ADB is more suitable for instances where you have a lot of processing ADB is more suitable for instances where you want that a particular action on a DB Table triggers a BW process. ADB adapter is best for publishing from database. For simple inserts and updates then ADB subscriber is best.
JDBC Palett:
In JDBC we can Query using JDBCQuery and manupulate data using JDBCUpdate. We can perform CRUD operations. JDBC palette for both run-time and design-time operates in the same way. No Trigger – BW JDBC does not have a mechanism to detect database changes in a push environment. In case of insert or update to database then check if you have complex JDBC inserts, transaction management and other dynamic queries then JDBC activities are best. JDBC is more suitable for running dynamic code where in runtime you can execute statements with different values depending on process execution.
Difference :
1)The advantage of ADB is that it has exception tables that will record any insert or update errors.
2) once it exits the BW process based on the transport used it can guarantee that a modification is delivered using the RV or JMS transport framework to handle that guarantee.
3) ADB lets you do upserts while JDBC you will have to check first before doing an insert or update.
4) JDBC is often alot easier to configure and is more direct to the datbase. ADB you are abstracting the DB to an adapter.
5) ADB Adapter will generate a schema for you to match the table you are trying to write to. JDBC you will just write your own SQL statements.
6) You really don't need to know SQL using ADB (although it is recommended that you understand it and as a developer you should know it) but with JDBC you have to know SQL.
No comments:
Post a Comment