DSXchange

WebSphere is a product name (marketing term) that IBM uses to group a whole range of middleware products. There's nothing to learn other than that fact, which will help you to find DataStage-related information on IBM web sites.

Tedious to reload the 20 million rows, however.

This suggests you're not getting detail lines on the other output link. This, in turn, suggests that your schema for the other output link is not correct for the detail lines. Get that right first. Then use a Modify stage on the rejects link to convert the raw row to a string, and decompose that str...

A hashed file is not intended to be opened for viewing by humans; there are lots of pointers and other stuff intermingled with the data. If you do open a hashed file with a text editor, make sure you quit without saving; otherwise you will corrupt the hashed file. Guaranteed. The hashed file is wher...

Is your code based on the way that job sequences do it? They seem to work OK - could be worth inspecting their code!

Do you mean a sequential file from which you're reading, or to which you're writing? You can not avoid anything in a sequential file from which you're reading; the very nature of a sequential file is that you have to read past every byte to get to the next. So you either have to pre-process the text...

You will need to be much more precise. Most DLLs contain multiple entrypoints, exposing more than one function. The individual functions are callable. It may be that your particular DLL contains only one function, but you did not say so. It's not easy to call them, but you can set the functions up a...

Where can I find some information about that functions dedicated to multivalued fields? DataStage BASIC manual, Chapter 2, especially the section on Dynamic Array Operations - multi-valued fields are a specific instance of dynamic arrays. There is a table of the vector functions in this section - m...

Assuming you have a VOC pointer to your hashed file (search for SETFILE), the command is This you could also have found with a search.

Could be a timing issue. Try a short SLEEP or NAP after DSDetachJob().

The return value from DSDetachJob() is its status. Check to see whether its value is DSJE.NOERROR or something else. If something else, screech to a halt and figure out what's happening.

DS_METADATA is a dynamic hashed file in which Table Definitions are stored. The directory should contain exactly three files, called DATA.30, OVER.30 and .Type30. If this is not the case, then you have major problem in your repository. In particular, DS_METADATA is not referenced by MetaStage. It ha...

Force the leading zeroes (Derivation) when extracting. Then you should be able to nest string_from_date in date_from_string in a Modify stage.

Yes and no. The IPC stage buffers data so that the next process (or next stage in the same process) can pick it up. Pipeline parallelism in parallel jobs is much more complete. Do you understand the relationship between stages and Orchestrate operators? Essentially each stage generates an operator. ...

You don't need a shell script - that was pointing out that Aditya's solution was not appropriate in a parallel job. If you follow the remainder of my advice, you read the file using a Sequential File stage. In parallel jobs a Sequential File stage can have two output links, the one containing valid ...

You appear to have a one-line parameter file containing ten comma-delimited values with which you want to replace eleven job parameters. The IsNull functions in your original code thwart this intent. Assuming you're using version 7.5, you can create a job sequence that begins with an Execute Command...