DSXchange

I have a sequence calling another sequence that calls 4 jobs (with dependencies). Both the sequences are set up as restartable. When one of the jobs fails, both the sequences have the 'Aborted/Restartable' state. When I fix the issue with the aborted job and just Run the outer sequence, it seems to ...

Yes, this is the error message when i try to compile..
"Link Collector stage does not support in-process active-to-active inputs or outputs"

I tried ORA->XFM->LC->XFM->ORA - this didnt work either.. (ofcourse with multiple sets of ORA->XFM as sources into the LC)

Should one of the inputs to a link collector always be a File? I am trying to read multiple oracle tables and use the collector (sort of like a funnel/UNION ALL) and load them to a target oracle table..

Thank you Ray and Craig..

When I import a table definition from SQL Server 2008 (via ODBC Table definition import wizard in DS), any column that is defined as Integer in the DB gets imported as Int but with a precision of 10 - anyone know where this default is set and if that can be modified?

Thanks in advance

Thank you - I just tried it and it works!! Did a 'google' on this topic and discovered that there's an article on IT tool box as well on this with the same solution. I still wonder if there was/is another way to do it from the Administrator GUI other than directly modifying DSParams I mean.. but for...

Thanks ArndW, but since there is a long list of params and eye-sore to see them, my question was how to organize/categorize them - there is surely a way, coz i see it done in other projects... you'd see a tree that looks like this (in administrator) User Defined l_ DB_Params (and the right frame wou...

How to categorize user defined environment variables? I have about 40+ variables and when I create them in administrator it looks cluttery and I want to be able to organize them, file related params, DB sources related and so on. I have seen that done in some earlier projects, but cannot see how to ...

it could be due to data issues, not necessary volume. In the director log you should see a pointer to the sqlldr files on scratch directory where the ctl and data file are landed during bulk load. If you see those logs, you should be able to tell what the errors are.

It can be achieved with a job sequence. Have the first job do the count and based on that result you can call the appropriate job - one to insert the dummy row (if count happens to be 0) or the job to do whatever it needs to do (if count > 0)

you should probably try changing the SQL in the oracle stage as

SELECT CAST(SUM(AMOUNT) AS DECIMAL(17,2)) from table

and try..
i would also add a nvl to the AMOUNT for better null handling..


SELECT CAST(SUM(NVL(AMOUNT,0)) AS DECIMAL(17,2)) from table

Can this be done with just a SQL, provided you are able to redesign the child table that has the rule expressions ? (also not sure if you have any OR or ! conditions.. but if all are AND expressions the table can be redesigned or staged to look like this for easy parsing and transformation Example: ...

Are you able to use a CDC change instead to determine if its a 'insert' or 'update' and then direct them to the appropriate target oracle stage ?

That's exactly my thought Craig and that is why i am surprised why a sequence cannot be called. However the 'ISD enabled' option seems to be available only in the jobs (job properties) and not sequences. Strange - but i wondered if there are work arounds ..

It looks like we can attach a DS job as the information provider, but can we attach a Job sequence? If a web service when called, needs to run 2 jobs in a row with some notification activity as well (for exception handling) need a sequence and cannot be accomplished with one job. Reading the manual,...