Multi Source execution plan build issue

Question

Hi,
I am trying to create/build a multi source (homogenous) execution plan in DAC from 2 containers for 2 of the same subject areas (Financial Payables) from 2 different EBS sources. I successfully built each individual subject area in each container and ran an executin plan for each individually and it works fine. Now I am trying to create 1 execution plan based on the DAC steps. So far I have done the following:
- Configured all items for both subject areas in the DAC (tables, tasks, task groups, indexesconnections, informatica logical, physical folders, etc)
- Assigned EBS system A a sifferent priority than EBS system B under Setup->Physical Data Sources
- I noticed that I have to change the Physical Folder priorities for each informatica folder (SDE for Container A versus SDE for Container B). I assigned system A a higher priority
- Build the Execution plan..
I am able to build the execution plan successfully but I have the following issues/questions:
1) I assumed that by doing the steps above..it will ONLY execute the extract (SDE) for BOTH system containers but only have ONE Load (SIL and PLP)..and I do see that the SDEs are ok..but I see the SIL for Inder Row in RunTABLE running for BOTH...why is this? When I run the EP...I get an unique constraint index error..since its entering two records for each system. Isnt the DAC suppose to include only one instance of this task?
2) When I build the execution PLAN, it is including the SILOS and PLP tasks from BOTH source system containers (SILOS and PLP folders exist in both containers)...why is this? I thought that there is only one set of LOAD tasks and only SDE are run for each (as this is a homogenous load).
3) What exactly does the Physical folder Priority do? How is this different than the Source System priority? When I have a multi source execution plan, do I need to assign physical folder priorites to just the SDE folders?
4) When we run a multi source execution plan, after the first full load, can we somehow allow Incremental loads only from Container A subject area? Basically, I dont want to load incrementally from source sytem A after the first full load.
4) Do I have to set a DELAY? In my case, my systems are both in the same time zone..so I assume I can leave this DELAY option blank. Is that correct?
Thanks in advance
Edited by: 848613 on May 26, 2011 7:32 AM
Edited by: 848613 on May 26, 2011 12:24 PM

Answer 1

Hi
you are having 2 sources like Ora11510 and OraR1211 so you will be having 2 DAC containers
You need these below mandatory changes
for your issue
+++++++++++++++++++++++++++++++++
Message: Database errors occurred:
ORA-00001: unique constraint (XXDBD_OBAW.W_ETL_RUN_S_U2) violated while inserting into W_ETL_RUN_S
You need to Inactivate 2 tasks in R12 container.
#1 Load Row into Run Table
#2 Update Row into Run Table
+++++++++++++++++++++++++++++++++
There are other tasks that has to be executed only once
(ie Inactivate the Below in One of the container)
SIL_TimeOfDayDimension
SIL_DayDimension_GenerateSeed
SIL_DayDimension_CleanSeed
SIL_TimeOfDayDimension
SIL_CurrencyTypes
SIL_Stage_GroupAccountNumberDimension_FinStatementItem
SIL_ListOfValuesGeneral_Unspecified
PLP_StatusDimension_Load_StaticValues
SIL_TimeDimension_CalConfig
SIL_GlobalCurrencyGeneral_Update <dont Inactivate this> <check for any issues while running>
Update Parameters <dont Inactivate this> <check for any issues while running>
+++++++++++++++++++++++++++++++++++
Task :SDE_ORA_EmployeeDimension_Addresses
Unique Index Failure on "W_EMP_D_ADR_TMP_U1"
As you are load from 11.5.10 & R12 , for certain data which is common across the systems the ETL index creation Fails.
Customize the Index Creation in DAC with another unique columns (data_source_numID).
++++++++++++++++++++++++++++++++++++
Task :SDE_ORA_GeoCountryDimension
Unique Index Failure on "W_GEO_COUNTRY_DS_P1 " As you are loading from 11.5.10 & R12 , for certain data which is common across the systems the ETL index creation Fails.
Option1) Customize the Index Creation in DAC with another unique columns (data_source_numID)
++++++++++++++++++++++++++++++++++
This changes were mandate
Regards,
Kumar