DSXchange

What happens to performance when you put in a sort stage and sort on COL_A and COL_B (hash partition on COL_A)?

You can use a batch or UNIX script around the various "dsjob" options. You can get a list of projects, then use that to get a list of jobs per project. Then use those lists to get the current status, then for those jobs with status running get the job logs. All using "dsjob" and ...

There is no general answer to whether or not to sort in DataStage or the DB. The only good answer is to empirically test it and choose the faster option. If the DB Server is heavily loaded and the DataStage machine is well-dimensioned with fast disks and memory it might well be much more efficient t...

What I've seen happen several times is that the DB2 table contains some non-nullable column which isn't declared or used in the DataStage job. Could that be the case here?

try Note that one cannot trim a Char(2) field, thus the comparison on 2 spaces.

No, you do not need to make any changes to shared containers between the two versions (assuming you aren't doing anything very esoteric within the shared container).

Are you doing pure inserts or upserts/updates as well? Can you turn on rejects from that stage and see what the SQL error code is?

It is a bit difficult in this case to localize the error since the full code is not there and the cause might be in those parts removed.

I usually go to the director, in the menu bar go to View and turn off the "show folders" then sort the list on "status". Note that this can take a while in a big project and you should turn on the "show folders" when you are finished to improve director performance.

If you have the 8.5 Software, then all you need to do is look at the PDF documents that delivered with it for the instructions.

You can't expect the forum members to guess your table name, column names, and data contents. Or the database type and stage you are using and any or all of those will play a role. In addition, the complete error message would also help in helping you.

Please cut-and-paste the error message and post it to the thread in order to make identifying your problem and solving it easier.

I don't understand some of the difficulties here. Can you, from the CMD prompt, execute your batch (without the CLS so you can see the messages) and get your job running?

Did you do this a month ago when you originally posted the question or is this a recent output? From this output I would look at process 4714516.

If you don't have write permission to the default configuration file directory, save the 1-node file anywhere on the machine (i.e. "/tmp/1node.cfg") and then reference that location in your runtime parameters.