DSXchange

Not unless you happen to have a copy of the source code for the osh executable (which I doubt). If it's any consolation, no-one posting here can determine a priori what the optimizer will do. The only one I know for sure is that it optimizes out a "do nothing" stage. Everything else is governed by t...

Where's it stored? My choice would be to use the SDK key management routines, and to use an appropriate command to reset the named sequence in the SDKSequences hashed file.

Always in parallel jobs add $APT_CONFIG_FILE as a job parameter with $PROJDEF as the default value of the job parameter. If your configuration file includes node pools and/or disk pools, individual stages can be constrained to execute in these. In an SMP environment there's not a great deal of valu...

For dsjob: $DSHOME/bin or, if you don't have DSHOME set, `cat /.dshome`/bin For dscmdexport and dsexport: C:\Program Files\Ascential\DataStage (or wherever else you installed the DataStage client. This answer is for versions 7.5 and earlier; I am not certain where the version 8 client installs by de...

nkreddy_3 wrote:... it's very critical for me to resolve ASAP.

OK, off you go then. If you have to resolve it, you have to resolve it.

One can only presume if you got the answer from here you'd claim credit for resolving it anyway.

No it doesn't. As with a do-nothing Copy stage it is simply optimized out of the score completely. Such a Copy stage can not display row counts on its output link in Designer (because it's not really in the job). Composing the score is, among other things, an optimizing and resource allocation proce...

So would I. You may find it easier to maintain if you store the output of your file-reading command in a user variable (in a User Variables) activity.

What does db2cmd do? It finishes, then there's a db2 command. What does it do? Then there's another two db2 commands. Is that really what you intended?

There is no regular expression support in DataStage server jobs. It has quite a powerful pattern matching capability, but based on character types rather than on values. The Matches operator and, to a lesser degree, the Index() function will prove useful. Of course you can write your own routine. Yo...

A Search of the forum would have found the dscmdexport and dsexport commands.

The third position in a Modify stage is actually 2 because, in Modify stage, everything uses C language conventions, including zero-based counting. However, in a Transformer stage, one-based counting is used. Further, the square-bracket notation for substring is not supported in the Modify stage; in...

IBM's International Classes for Unicode, or ICU , is the mechanism via which parallel jobs handle multiple character sets. You can also get information from the IBM website The xyz_cs.txt files are the names of the character sets that are supported by the xyz stage. Clearly your Teradata character s...

Define "performance" in an ETL context.

It's impossible to recommend techniques for "better performance" if one has no idea what "performance" is.

Also, there is no "one size fits all" solution. It always depends.

Two outputs from Transformer with constraints qty1=qty1 (or simply unconstrained) and the other qty1<>qty2. Bring them back together with a Funnel stage if needed.

Are there very many columns? If not, why not construct the entire line (with required embedded tabs only) in a single string field? Initialize a stage variable (let's call it svTab) to Char(9) and use svTab in any expression where you need a tab.