DSXchange

Use a database that does support transaction logging. This is a standard message from Informix when you use a database created with -nologging option. DataStage's philosophy is to log all messages from the database, on the grounds that the database vendor thought it important enough to inform you, s...

Well done! And the work you expended in your research will benefit you by establishing the neuronal paths that help you to retain that knowledge.

It may be possible. However, if you have use any of the activities that are new in 7.5 (StartLoop, EndLoop, UserVariables, Terminator), then it will not work.

If you import the table definition from your COBOL FD you will find that half your work has already been done; the data are already presented as SQL data types. The only thing is, you must have a 01 level item in the FD.

There is no log created during importation.

There is no log created during compilation, unless done via Multi-Job Compile.

Link collectors from active stages should must have IPCrow buffering turned on in the tunables.

IPC is a stage type, that makes row buffering explicit. However, both the Link Partitioner and Link Collector stages also use the same libraries to refer to the data buffers.

Try it without the trailing slash in the pathname.


I take it that the job does have a parameter called "Path"? Case and spelling are important.

My guess is that it's neither sensible nor possible. If you could store Word documents in a TEXT field, you could read WORD documents in a text editor. You can probably store Word documents in BLOB fields, but DataStage does not support this data type. You could certainly try using a Folder stage (w...

The most likely candidates would be the control and data files for the multiload operation.

Are your Teradata environment variables set up correctly? Since you have used the Teradata Enterprise stage, the answer is probably yes, but it's worth checking.

The Teradata Enterprise stage reads the results of the query into a virtual Data Set against which the lookups are performed. If the query returns no rows, then the Data Set is empty, and there's no way any lookup would work, hence the "no rows will be processed" message. Since we have no informatio...

Please provide solution ...... Providing solutions is the business that some of us are in, in exchange for reward. There are many solutions, which a search will reveal. But not matter which method you choose, you will have to tell DataStage what you want it to do. DataStage writes entries into job ...

What Vincent said also applies to performing a lookup against a Lookup File Set stage. I acknowledge that this is not relevant to the original post, but it's a "gotcha" in both situations.

You must comply with the rules. Change your load rule for the parent table so that rows are updated (rather than replaced), or allow automatic cascade on delete (ALTER TABLE), or drop/disable constraints before loading. Each of these is dangerous.

How do you specify the same remote file from a UNIX prompt? If you can do it from there (someone has mounted its file system), then you use exactly the same technique in DataStage.

Then, yes, it's true within the Lookup stage. But you could, of course, precede the Lookup stage with a Column Generator stage that uses a cycle from N to N for the column in question (one way to generate a constant).