DSXchange

Welcome aboard. :D
The short answer is yes. Specify delimited format when you're reading, and fixed width format when you're writing.

The job log is not a file. Get this idea out of your head. The job log is a sequence of records in a database table (RT_LOGnnn) in the Repository database (project schema). Therefore you view it using SQL. They have also provided some API functions, and exposed these as options to the dsjob command.

Is the variable EMAILSUBJECT actually defined? If it were not, you would see the behaviour you are describing.

Second set. 1) Hashed files can not be used in parallel jobs. 2) Why do you insist upon discarding the benefits of parallelism? 3) Pipeline is automatic. Inspect the generated OSH. 4) How long is a piece of string? 5) Throw away the server job design and the mindset that accompanies it.

Sure, just go ahead and load.

But it will be faster without the constraints. If you can promise the DBA that your data will not violate any constraints, then it ought to be OK to drop them and later re-impose them. After all, an empty table is not needful of constraints.

Iconv() is a fairly clever function. Try

Give us your thoughts. In a server job you would have looked at Ereplace() but this is not available in parallel jobs. Clearly an approach that uses substrings and concatenation, plus If..Then..Else, seems to be indicated. Can you guarantee that the string-to-be-replaced occurs not more than once in...

1) Lookup File Set stage if the data volume on the reference link will fit in memory, otherwise consider Join or Merge stage. Lookups are performed with a Lookup stage, not with a Transformer stage. 2) It's largely automatic. Use (Auto) unless you have a specific need to partition on a grouping or l...

DataStage clients use port number 31538; if this were disabled you would have been getting "connection refused". Even so, it might be worth checking that this port is open through any firewall that you have, and that it is properly recorded in the /etc/services file.

"Project" UV is the DataStage installation directory. It's called that for historical reasons. It's the initial attach point for the Administrator client. When you telnet, and get to the ">" prompt, you can execute the command LOGTO UV - having done that, can you then execute SELECT * FROM UV.ACCOUN...

Can't reproduce it. Can you generate a job report (click the "Internet Explorer" icon second from the right in the Designer toolbar) and publish this on a web site somewhere so we can look at it? At worst, publish all but the bitmap here. If you could also provide the 10 rows of data that would also...

You can always send to an email address that is a "list" in your email server.

Doing it from a routine incurs one of two fairly severe penalties. Either the routine executes a shell script (with all the overheads of establishing and managing a new shell and return structure), or the routine uses BCI functions (which involves paying for a Data Direct ODBC licence). Why re-inven...

Welcome aboard. :D The list is in the GUI itself. Every site is different; you will have added your own Transforms, Routines, Data Elements and so on. So you are better off browsing through the Repository where you are working, inspecting those that look interesting, and asking your colleagues for h...

That message has not come from a job sequence. It has come from a job that contains a Transformer stage. You need to track that down to figure out where the problem arose.