I want to create views in greenplum HAWQ using a simple talent task that will basically have a file containing all the views, then I need to run a CREATE VIEW script.
Since these representations (50-60,000) come from the oracle system, I need to find those that we could not create.
Here is the layout for my problem: 
I have a view already in the database, and I want to create it 3 more times. This obviously will not work.
Here's the conclusion:
Exception in component tGreenplumRow_2 org.postgresql.util.PSQLException: ERROR: relation "ad_apps_dependencies" already exists at org.postgresql.core.v3.QueryExecutorImpl.receiveErrorResponse(QueryExecutorImpl.java:1592) .--------------. | tLogRow_4 | |=------------=| |componenterror| |=------------=| |componenterror| '--------------' .-----------------------------------------------------+-------------------------------------------------------------. | tLogRow_5 | |=----------------------------------------------------+------------------------------------------------------------=| |result |result1 | |=----------------------------------------------------+------------------------------------------------------------=| |ERROR: relation "ad_apps_dependencies" already exists|CREATE VIEW SYSTEM.AD_APPS AS SELECT * FROM APPLSYS.AD_APPS| '-----------------------------------------------------+-------------------------------------------------------------' .------------. | tLogRow_6 | |=----------=| |subjobError | |=----------=| |Subjob Error| '------------'
I want to get this output 3 times since I am trying to run a query 3 times. (In the final version, I would create another file containing only those requests that failed, so we can fix them later, but this is a very important point.)
As a workaround: I could move this part of tRow -> OnError -> FixedFlow -> FailedViews to another job, but this is not an elegant solution.
greenplum etl talend hawq
Balazs gunics
source share