DSXchange

kumar_s wrote:If both the stream keys are varchar, shouldnt it take care by it self?

no .. it can only handle variable length values

like varchar 10 can handle "abc" "abc1234567" .. but it still treats "abc " different from the way it treats "abc"

The problem is... data is getting rejected at transformer stage .and i am getting warning message as ...[APT_CombinedOperatorController(1),0: Field 'PROCESS_DATETIME' from input dataset '0' is NULL. Record dropped.] this is different problem .. and covered well on DSX!! this happens if you are doin...

Yes they are part of the primary key and i have handled the same properly in my query itself. it could be that you need to modify your query a bit .. lets say your primary key is c1+c2+c3 and your unique index has c2 in its definition: with the logic you have, DS will check if c1+c2+c3 in the incom...

search on DS Help resulted in the following function

-----
TrimLeadingTrailing
Returns a string with leading and trailing whitespace removed
string (string)
result (string)
-----

I remember using this in XFM .. you can try and see if its available in Modify stage as well??

the unique key columns mentioned in 'rep_fin_asset_UIX1' index definition are part of the key columns in the query??

I understand. But if you change the target to Char. It can handle decimal values. I am not too sure, but give it a shot. Target is already CHAR and it is Fixed Width file . hmm!! in that case I think you are stuck with implementing nulltovalue logic before writing to the file!! I havent seen any ge...

If I were developing this code in Server Edition, I'd simply use the ASCII() function. How can I do the same thing in Parallel Edition? parallel transform doesnt have ASCII function ..but the BASIC transform does. The simplest solution would be , if you can afford the performance loss, using this s...

did you check the INSERT sql and UPDATE sql code that was generated and if that has correct key settings??

Update then Insert--Upsert mode-- will only try to update the record if it finds the duplicate key combination in the table otherwise it will happily tries to insert the record!!

gsherry1 wrote: Then after splitting by rec_id, land the file and reread with more specific schema to each record. Is this what others are doing?

Yes .. we have similar requirement and we followed similar technique

I had similar problem. In my case, the isnull check and the transformation logic is in StageVariable derivation. If I move that to output column derivation, DS is happy!!

Thats one of the things I have actually looked at already .. It did not work in this case .. as I have mentioned DS tried to enable this FK first and determined that corresponding PK is not yet enabled so threw a warning and continued with rest of the constraints.. I think, Defering works when we ar...

I have solved this problem. It has to do with the order in which Constraints are enabled .. We have given name to Primary Key (on Col1 in Eg) which comes later, in alphabatical order, than the reference_key name in question. So DS tried to enable FK but by that time the PK is not yet enabled. If I r...

did u check the offending record using exceptions clause alter table <table_name> enable constraint <Foriegn_key_constraint> exceptions into <exception_table> [/code] straight run or ALTER command @sqlplus prompt with out the exception clause is not at all complaining .. but only when I do it from ...

The error is about the exixtance of the parent column. Check whether the give name is correct. If not try to have table.column name in the statement. thanks Kumar .. BUT as I have said same alter command worked PERFECT when I ran it from SQL promt (sqlplus) .. I will try to alter the constraint def...

Options: Write Method Load Write Mode Append Disable Constraints TRUE -------- Table Structure Col1 Non-Nullable PK Col2 Nullable FK references Col1 in the same table .. .. There are some more referential intigrity constraints and CHECK constraints on this table .. also some indexes were defined ---...