DSXchange

Oracle enterprise stage is failing to connect to the database if the password has a $ character in it. I tried "$DS_ENABLE_RESERVED_CHAR_CONVERT = True" hoping that this column level setting would also work for the password. But that did not work. I am setting the user id, password for the...

Typically comp is binary for integer. Try using "DataTypePicComp". In version 6.0 it used to work only for int types (max 4 bytes). May be the limitation is still there. If your input is 8 bytes long you might consider using "DataTypePicComp2".

Can you check if it works if the target is a dataset or a file set instead.
If yes, the problem may be in the data and that needs to be taken care of in the transformer.

Oracle synonyms are at table level not at column level. So that option is not available.

Custom SQL?
I am talking about direct load option in OracleEnterprise stage in 7.5 parallel. Are you talking about a different stage?
I 've chosen the Write method=Load and Write mode=Append. And with this setting I am unable to see any way to enter custom SQL. Am I missing something?

Thanks
Rati

Custom SQL?
I am talking about direct load option in OracleEnterprise stage in 7.5 parallel. Are you talking about a different stage?
I 've chosen the Write method=Load and Write mode=Append. And with this setting I am unable to see any way to enter custom SQL. Am I missing something?

Thanks
Rati

How do I do that in the oracle enterprise stage. Is there any option setting that allows me to put quotes in the column names?

Thanks
Rati

i am tying to load a table via dirct load. This table has year and month column. While running the job sqlloader abends with invalid keyword error message (as I have columns named as year and month). I can write a custome sqlloader control file to load the data by putting quotes around year adnd mon...

I use ODBC stage to read ref cursors returned from oracle stored procedures.

You should have only one output parameter for the ref cursor.
Define the columns that you expect to get from the cursor.

and you are good to go.

if position and length of your file_name field in your lookup file is fixed, then you can build a has file.

then use the hash file in the transformer to do the existence check.

Take a look at transform "RowProcCompareWithPreviousValue". You will need two stage vars. Input shhould be sorted on table name. Output should be a hash to hold the last record for the table as the transformer stage would be outputting records in the following manner tab1,col1 tab1,col1,co...

Dos this file has a header or trailer record that has the record count in the file? If yes, you may think of writing a shell script to check the total bytes of the file with the computed total bytes for size 100 and for size 101. And see which one matches. Then call the fix_100 or fix_101 job to do ...

Typically I do it using a server job. The OCI stage has the "generate ddl option" which prepares end excutes the DDL.

Show the simple job would be
CFF => TRansformer => OCI.

And to run the job I create empty file using touch filePathName.

Hi Kash,

Did you finally manage to get it working? I've the same requirement and I've tried lots of combinations, but nothing seems to work!

Thanks
Rati

Thanks Kim and Ray. I tried pritning upto @RECORD<9> But nothing showed up. The OLE type I am looking for is CCustomInput and CCustomOutput. I guess the OCI stages have those OLE type. (I may be wrong). How do I know whether a particular record element for ex @RECORD<7> is multi-dimensional (From yo...