DSXchange

This is not a DataStage problem. You should probably show your error message to your DB2 DBA.

I had this problem once, but it was resolved by the DB2 team.

I would recommend making a VarChar stage variable "svCAT" derived by "Link_Inv_Retraite_VCN.CATEGORIES[1,2]:','" then the constraint: NOT(IsNull(Link_Inv_Retraite_VCN.ACT_INITIAL) OR IsNull(Link_Inv_Retraite_VCN.LISS_FINAL)) AND INDEX('07,15,01,16,02,',svCAT,1) This uses the IND...

You can also use the field function which split the string as all of you CATEGORIES starts with a number then a delimiter and a value. Not(IsNull(Link_Inv_Retraite_VCN.ACT_INITIAL)) And Not(IsNull(Link_Inv_Retraite_VCN.LISS_FINAL)) And ( Field(Link_Inv_Retraite_VCN.CATEGORIES,"-",0) ="...

If you specify 0 or a negative value as start argument, the value is assumed to be 1. Try changing you constraint for this one : Not(IsNull(Link_Inv_Retraite_VCN.ACT_INITIAL)) And Not(IsNull(Link_Inv_Retraite_VCN.LISS_FINAL)) And ( Link_Inv_Retraite_VCN.CATEGORIES[0,2] = "07" Or Link_Inv_R...

Not(IsNull(Link_Inv_Retraite_VCN.ACT_INITIAL)) And Not(IsNull(Link_Inv_Retraite_VCN.LISS_FINAL)) And (Link_Inv_Retraite_VCN.CATEGORIES[0,1]='7' Or Link_Inv_Retraite_VCN.CATEGORIES[0,2]='15') this should work if the character you are comparing is the first one in the string, otherwise, change start f...

The problem is solved. Even now, I don't know the DB2 NLS characterset as the JDE team couldn't give it to me (sik), I guessed it was NLS as it allowed chinese, japanese, czech characters. As I wrote, My DataStage NLS was UTF8. After some exchanges with the Jd Edwards ERP Team it seems this version ...

Can't anyone help me ?

Please say me if the way I explained my problem isn't clear.

Hi, I have a SQL server 2000 source database (MS1252) with czech characters in Nvarchar (Unicode) fields. My target database is a DB2 database (JD Edwards) which allows czech characters. Here is the global design of the job (simplified) : Job1 : ODBC stage on the sql server database (NLS MS1252 with...

a) Write good records to new table. b) Drop old table. c) Rename new table to old table. That's a good solution if you don't have constraints like foreign key pointing on this table. But I agree with you, a job is not needed for this kind of task. rafidwh, why don't you use SQL ? It's much more fas...

Hi,

You wrote the Jobs are running with the old Database. What's the difference between your databases ? Version ?

Have you tried using an oracle Oci Stage instead of an Oracle Oci 8 ?

I just don't understand why I can't import chinese characters in my Oracle Database . The NLS_characterset is UTF8 and all the varchar column are defined as varchar2(X Char), so stored in char mode and not byte. Ray Said : Almost everything is DataStage is based on characters, not bytes My Oracle St...

Hi, I have the same King of problem. gpaton wrote : If you were to reimport the Database meta data, you will find that for Char and Varchar columns, the length will be a multiple of the byte length ( depending upon the value set in the DataBase ). The soulution is to multiply the length values in th...

I just don't know why there is a select.

As I thought it was suspicious, I checked the option again and it's "Update existing rows only". And How could it be a "Select", there is just an Input Stream connected to my ODBC stage.

Hi, I am using an ODBC Stage to access a SQL Server 2000 Database and updates rows on a table. The option is "Update existing rows only". My Job had no problem but i added a column in the ODBC stage and now, There's an error on the SQLPrepare Select which determines if the rows exist or no...

Hi, I use DataStage Server 7.5. I use "Shared Container" with parameters such as the database table. The job is supposed to update some rows of the table but sometimes it updates 0 row. I don't want to use a "Look Up" to know if the row i want to update exists in the table becaus...