DSXchange

... and it all changes in the next release, when a common (open) repository will be implemented.

If you want "" specify this rather than the final ELSE @NULL. Otherwise you've pretty much got an optimal expression. Since it does not perform arithmetic, precision does not come into the picture.

I believe that there is a maximum of 127 output links, but would have to check.

Nothing to stop you modifying, or creating your own, dsjob command. The API is documented ( Server Job Developer's Guide and Parallel Job Developer's Guide ). I think it would need a whole lot more checking to be safe than just who is executing it. Maybe only allow locks held by defunct processes to...

In short, check everything to do with the connectivity path. Thoroughly.

Check it again.

Get fresh eyes to look at it - you may be so familiar with it that you're missing seeing a vital "." or similar.

Looks like the network has dropped out around about when DataStage is trying to close the connection to Informix.

Does this happen reliably, or only once? Usually a dedicated network segment, isolated from the remainder of the network, is a good idea in these circumstances.

You can use a UV stage to access the hashed file. Declare only col1 to be Key; declare counter not to be Key. In the Transformer stage check the "multi-row return" check box. This will not be fast. To get some speed you would need to create an index on col1 in the hashed file. There is no reason to ...

Sounds like /tmp is full, or maybe you don't have write permission to /tmp (an unusual way to configure a UNIX system).

There's nothing wrong with the design. It just can't be done with a hashed file (other than in its guise as a UV table). Accessing a hashed file via the Hashed File uses the key, the entire key. It is not equipped to use partial key searches. The entire key (that is, the physical representation of t...

... or using substring and concatenation, for example

Are they really VarChar in SQL Server, or NVarChar?

Use stage variables to "remember" the previous row, and the values therein to generate the value in each output row (governed, almost certainly, by some form of If .. Then .. Else expression).

Or run a pre-FTP command such as and use that file as the one to stream into your job.

You will read in other posts and in the DataStage BASIC manual that arithmetic is performed with the full precision of which the machine is capable, and the result rounded to four decimal places (by default) for display. The default can be overridden in routines by a PRECISION declaration. This is n...

Use a job sequence that has a Wait For File activity. That's where you set the timeout, and the file path. Have two outputs from this activity, one that is taken if the file appears, the other if the file does not appear (control this on the Triggers tab). Connect the second link to a Notification a...