DSXchange

IMPORT the table definition and use that. Chances are that you've missed one key column, or more than one.

How many columns are in the hashed file? How many of these are actually used in the job? Are you enabling read cache for the hashed file?

Now you DO have to write some code. You can write DataStage BASIC code, or you can write C code. The latter is harder to write but easier to call from the command line. The DataStage C API is documented in the Server Job Developer's Guide

Search the forum. Is the job really running, or are you relying upon what you see in Director (which is an entry in an RT_STATUSnnn table, and which may not have been able to be updated if the job aborted). Check to see whether the process ID still exists.

Not possible with a query. You need a routine to iterate through the job names in DS_JOBS, get the job number, retrieve the //PURGE.SETTINGS record from that job's log, and build your report in that way. In hindsight it is possible with a query; a UNION across all RT_LOGnnn tables. You'd probably ha...

Without checking, you probably need to restart DataStage services before the new value is picked up.

Increasing MAXFETCHBUFF will not affect existing jobs; only sufficient space is allocated to transmit each buffer (row).

You don't need to develop a routine, though it would be more flexible. ExecSH (or ExecSHSilent) is supplied with the product - all you have to supply is the command line for dsjob command to dump the log.

Benefit of the doubt for now. I suspect parallel, since the job type is marked thus - but, then, that's a forced choice!

General probably IS the appropriate forum, since the answer will be the same irrespective of job type. If you want the entire log, use an after-job subroutine. Perhaps use ExecSH to invoke dsjob, or "roll your own" using the DataStage API. Better, use a downstream activity in the job sequence, becau...

Make the "lookup failed" rule in your Lookup stage "Continue", and the fields derived from the reference input link Nullable. Downstream add a Modify or Transformer stage to handle the null, replacing it with -1. You are still thinking like a server job developer. In parallel jobs the rule (mainly) ...

000 is the ASCII decimal value for NUL. This is not the same thing as NULL, which is represented differently by different vendors. For example, within DataStage an out-of-band NULL is represnted by -128. 0000 is the Unicode decimal value for NUL. You can also specify non-printing delimiter or quote ...

(a) Yes (b) Yes to set them globally across all projects. Otherwise set in the Adminsitrator client for that particular project. Optionally load them as job parameters so you can override the values on a per-job basis. Different UNIX variants name the shared library search path environment variable ...

If DCount() does not work in the expression, use Count().

How - precisely - have you specified the reference to the job parameter in your DB2 API stage? Copy/paste the field value and post that. Also post the exact name of the job parameter from the job parameters grid in job properties. Usually when an error message contains parameter references that incl...

to determine exact version on server.

"Help > About" from a client will give both client and server versions.