What can cause to data collection "Star" job to fail?
Hi All,
We use Oracle g10 as our DB and recently we decided to configure the data warehouse.
All the data collection jobs run .o.k., only the "Star" job fails.
I saw in the log "jts-etl" an error message:
Java ETL (STAR): Dimension build
"com.ibm.team.datawarehouse.service.starjobs.internal.dimensionbuild.ResourceDimension" failed
Any ideas what can cause the "Star" job to fail ?
THX,
Ziv
7 answers
Francesco, thanks very much for your answer.
Exactly, these tables does not exist. My concern is WHY this tables does not exist!!?? The repotools -createWarehouse worked fine with no errors. Thus, I'm afraid to recreate the DW again and stumble on the same problem...
So yes, I will open a PMR to fix this issue...
Regards,
ISMAEL
Exactly, these tables does not exist. My concern is WHY this tables does not exist!!?? The repotools -createWarehouse worked fine with no errors. Thus, I'm afraid to recreate the DW again and stumble on the same problem...
So yes, I will open a PMR to fix this issue...
Regards,
ISMAEL
Hello Ismael,
i.e. in DB2: select * from SYSCAT.TABCONST where CONSTNAME = 'PROJECT_FK1' and tabschema ='RIDW'
so the table affected by the constraint is RIDW.D_PROJECT
Best Regards,
Francesco Chiossi
RIDW.PROJECT_FK1
is not expected to be a table, but a constraint.
i.e. in DB2: select * from SYSCAT.TABCONST where CONSTNAME = 'PROJECT_FK1' and tabschema ='RIDW'
| CONSTNAME | TABSCHEMA | TABNAME | OWNER | OWNERTYPE | TYPE | ENFORCED | CHECKEXISTINGDATA | ENABLEQUERYOPT | DEFINER | REMARKS |
| PROJECT_FK1 | RIDW | D_PROJECT | DB2ADMIN | U | F | Y | I | Y | DB2ADMIN |
|
so the table affected by the constraint is RIDW.D_PROJECT
Best Regards,
Francesco Chiossi
Hello Ismael,
RIDW.D_DATE is indeed expected to be populated at the time the data warehouse is first created, so I suspect there was a problem at that stage.
If it's a new Data Warehouse and (meaning you have no concerns on losing trending data) I would try dropping and recreating it (paying particular attention to the logs for this operation).
Best Regards,
Francesco Chiossi
RIDW.D_DATE is indeed expected to be populated at the time the data warehouse is first created, so I suspect there was a problem at that stage.
If it's a new Data Warehouse and (meaning you have no concerns on losing trending data) I would try dropping and recreating it (paying particular attention to the logs for this operation).
Best Regards,
Francesco Chiossi
Hi,
Did you finally fix this issue about STAR ETL Job? Last year I reported same error in the Work Item Defect 267123, with no answer.
Basically, the jts-etl.log shows the following error:
2014-02-27 09:25:50,909 [ WebContainer : 12] INFO ervice.starjobs.internal.StarRemoteSnapshotService - Java ETL : Started executing query: "com.ibm.team.datawarehouse.service.starjobs.internal.dimensionbuild.ProgramProjectDimension=>SELECT PJ.PROGRAM_ID AS PROGRAM_ID, PJ.PROJECT_ID, PJ.NAME, PC.NAME AS TYPE_NAME, PJ.ISSOFTDELETED FROM RIODS.PROJECT PJ JOIN RIODS.PROJECT_CLASSIFICATION PC ON PC.PROJECT_CLASS_ID=PJ.PROJECT_CLASS_ID WHERE PJ.PROJECT_ID>-1"
2014-02-27 09:25:50,985 [ WebContainer : 12] ERROR ervice.starjobs.internal.StarRemoteSnapshotService - ORA-02291: integrity constraint (RIDW.PROJECT_FK1) violated - parent key not found
java.sql.SQLIntegrityConstraintViolationException: ORA-02291: integrity constraint (RIDW.PROJECT_FK1) violated - parent key not found
Thanks in advance!
ISMAEL
Hello Ismael,
an error like "
This usually hints to data corruption in the data warehouse.
I would suggest opening a PMR to get this investigated in depth as it might be necessary to share some queries and data from your data warehouse.
Best Regards,
Francesco Chiossi
an error like "
integrity constraint (RIDW.PROJECT_FK1) violated - parent key not found
" sounds like there is a key that references another table, but in this table that value doesn't exist.
This usually hints to data corruption in the data warehouse.
I would suggest opening a PMR to get this investigated in depth as it might be necessary to share some queries and data from your data warehouse.
Best Regards,
Francesco Chiossi