DSXchange

Loading the T_ORDR_O into a hash file is not a bad idea, but if you try to load the 400million data into hashed file, you will be forced to operate in second half of the inverted 'U' curve in performance vs load. Still you need to explore, some of the option which might be more effecient in database...

Ray - Iam curious to know the rest. I guess Filter option will be one among it.

Or convert them into varchar datatype.

Prefer a sequential file stage if the files have the same structure - Got to see it now.

It appears to me that, I have seen these file, in some of my project directory, but Iam sure I dont have RTI component in any of it. I ll check it on Monday.

What is the job design? Which stage the warning is pointing to? Do you see any other related warning/Error in the job log?

Do a reset of this job, and check, if you find any "From Previous run..." log. Try to limit the number of rows for a run, and check if the job able succeed. If not aborted with various less number of set, you can try increasing some of the data type length or the disc space.

What is the job design? And how did you split it ?

Hope change in the requirement came as Change Request, rather than a raised as a bug. Do you have historical source data for re-load?
In case of CR the method of load can also be in Black and white.

Or filter option, if the file name are in same pattern.

I guess, those soa file are the temp files that are created during the sort operation. Try to move those file to a temp directory, to confirm the seamless operation of your jobs. Later delete it.

If you are looking at the different values of a variable in different user id, $ENV will be prompt aproach, as mentioned in the above.

Usiing dsjob comand, and by passing necessary parameters, you can fetch the number of projects available in the server. And by using the same dsjob and passing by different parametet, you can find the list of jobs available the project, and the respective time of finish can be found.

ORA-01458 invalid length inside variable character string Cause: An attempt was made to bind or define a variable character string with a buffer length less than the minimum requirement. Action: Increase the buffer size or use a different type. Is ther any way, the columns are been mapped wrongly i...

It is a 'nice to have' object. You need to build data elements using the existing standard, datatypes only. By creating this, making more understandalble format. You can built using basing data types as well.