DSXchange

shyju, I am still not sure what you are trying to do. Let's use the following example of input files; file1 has 6 columns, file2 has 4 and file3 has 3: File1: a,b,c,d,e,f a1,b1,c1,d1,e1,f1 a2,... File2: 1,2,3,4 1.1,1.2,1.3,1.4 2.1,... File3: x,y,z x1,y1,z1 x2,... What do you want your resultant sequ...

Let us take a step back, are you using this hash file as your main input or as a lookup? If it is your main input, then you need to use different syntax in your select.

Raj, when you read from a hash file that has multiple columns defined as keys you need to specify all of the keys to read a record. So in your case you are not getting anything because you have keys x and y but are accessing using only y . If you wish to use this type of a select on a hash file you ...

Look at your execution log file and see if there are messages in there about rejected records

pongal,

could you do us the favor of pasting the contents of your load log file located at worknode2/scratch/ora.9958.892635.0.log to this thread?

TheBird, the rebuilding was not his question, though. I think that part is clear - an unusable index needs to be rebuilt. But it was the use of DELETE/SELECT that he is asking about on a table with an unusable index. the Oracle setting that I gave will work at Oracle 10.x; I am not sure if it existe...

What about "skip_unusable_indexes true", that should allow the delete operations to run.

Can you disable that index and do an appropriate SELECT on the table?

gsherry1, manually modifying contents of .dsx or .xml export files might be complicated, but it can be done. Of course, making an error will result in the job either not importing, or not compiling/running after import. If you create a simple job wth few stages and even fewer columns, export it and ...

If you chose a "table load" (i.e. bulk load) then you will have an "index mode" option in the Oracle stage, set to either "maintenance" or "rebuild". I think that in the log file there would have been a warning from the stage stating something like "unable to build index '....' on table"; at least t...

How do you want your output file to look? one row from A, one from B, one from C and so on? all A's then B's then C's?

Yes, this is where column propagation works wonders in Px. You could also declere the sequential input with just the initial 3 columns and then click on the "suppress row truncation warnings" and ignore the rest of the input columns.

Yes, I think you've identified the cause.

But in my windows scheduler I can choose any user to execute the command as, so I am not sure if you can't change your system's access rights around to allow the DS-userid to also be fired off from AT

Hmmm... It is hard to give an answer here; how are you "merging" the 3 files with different metadata? This means are you applying changes so that each file changes the columns to some standard and then outputs to a common file {all records in A} then {all records in B} then {all record in C}? Or do ...

Eric, OK, I jumped the gun and thought that you were getting IDs mixed up. In that case, the User/Password that you enter in scheduling the job should be that which runs it, just as you did (correctly) state in your original post. So if the job isn't even getting started then either you have a type ...