DSXchange

Its a DB2 error, Check with your DBAs..it looks like your server name is incorrect.

I came across same error with the UDB stage, and to resolve this problem, I had inserted the date field in internal format and it worked fine for me. But after reading Craig's post, I think it might me the same case for DB2 where one need to add the time part for the date field. Let me play around a...

If you find the record in the Hash file (i.e. Update Action), you do not execute the routine..this you can achieve by creating two o/p links one for insert and one for update. If you want to do it in one link (Insert/Update option) then you need to write a derivation for key column..where you check ...

Then why this error?

What I can suggest is use, a different character and not blank for replacing null.

If you have to do it as after job, then its only possible if you include that in very last job of your sequence or else you can always use execute command stage in your sequencer for executing unix command.

If I am not wrong, in Oracle, it does equivalent of trim on the column and thats the reason you are getting this problem.

Try writing some character and not space, see if this works.

How you are trying to read the file? I think your problem is with the column delimiter...

How can you get the link info in the same job, with @INROWNUM<2..you need to have a seperate process for this, can be after job. You can write a custom routine for the same..use DS functions

I agree with Sainath, the best option would be to use a character, thats very unlikely to appear as data in your file. May be ~ or `

As far as you don't remove your link for target stage, your derivations in transformer won't change.

I don't have 7.5 as of now, but I think it should come with Version Control.

And are you trying to say that you will have all the environments in one box ???? Please tell me thats not right...

Whats your source?

Try using PIVOT stage...