DSXchange

The SQL issue has been covered in this forum; in addition it is very important for you to do a DS.TOOLS and reindex; the secondary indices are linked to by absolute pathnames, so doing an OS level move of the files and indices will cause corruption that requires index rebuild.

Lance - do single quotes around the default date value make a difference?

It is hard to compare directly; if you are reading from and writing to sequential files there might be a speed advantage on the "leaner" linux system. If you are accessing a local database then that manufacturer's implementation makes a big difference. If the database is remote then the networking p...

subramanya,

do you have the option to go through an ODBC stage to see if that works differently, and if it does work is it fast enough for your purposes (as a workaround)

mnvihol,

Sorry, I said "source" but meant "sequential" stage; but the method is the same in both places.

you need to click on one of the columns, then RIGHT-CLICK and choose EDIT ROW.

Either that or look it up in the documentation.

The compiled server code is portable. There used to be a FNUXI type issue between some platforms, but I recall reading here that this has been solved; so you should have no problem porting the code. Although the issues around big-endian and little-endian are supposedly solved, I think that if I were...

dls / Fernando,

before you take the class, the documentation does detail the procedure and that might give you a broad overview; but I would highly recommend taking the training before doing any work.

gsherry1, you can use the DSGetJobInfo() routine to get information about stages and links in your job (using key DSJ.ME or other jobs by doing a DSAttachJob() to them), then use DSGetLinkInfo() and DSGetStageInfo() to get the information you are seeking. Note that these are informational routines, ...

This has been dealt with before, you need to edit the row properties for the columns (go into your source stage and right-click on a column then edit-row). You can specify attributes about whether or not to fill nulls, what the fill char is and what the fill length is for nullable fields.

Jack,

this is the name you want to give your output file; browse just lists all the files in the folder chosen. It is best not to overwrite existing files, so choose your own name and let the job run.

Go into the Administrator, and for in that project properties check the box "enable administration in the Director"

Call JobHandle = DSPrepareJob(JobHandle) to do this.

I don't know. Perhaps when you remove a link the data that is causing the error doesn't get writtine to Oracle. What kind of an Oracle stage are you using? Do you have NLS enabled? Do you have "unicode" set in the varchar/char column attributes?

The DataStage command "LIST.READU EVERY" will show, in the "Active Record Locks" at the end, an ID which also contains the workstation/PC name at the end. If you have no other users on the system then this list should be short and might show the offending process. You should get your DataStage admin...

Kumar, only use the UNIX kills as a last resort. The preferred interface is in the Director -> Job -> Cleanup Resources. You can often tell which pc-name has a job open, using the lock information. If the job is not "running" and you can't edit it then probably another designer session has it open -...