DSXchange

Have you considered the implications of removing the delimiter character from a delimited file? How are you employing the cat command? Through an Execute Command activity presumably. In that case you can pipe the output of cat through one or more tr commands to remove or translate any character in t...

Is exactly that schema used in the generated OSH? Did you use an automatically generated Filler? (There may be a bug in this functionality.) The schema file you posted looks OK to me.

Most passive stages, the CFF included, have only one output link. There may also be an optional rejects link. A Copy stage copies its input to one or more outputs. It can drop or rename columns as part of this process. A Data Set doesn't "work" in any sense you're asking about; it's a location in me...

The precise length is the width of the column in which the decimal number is stored. This can be ascertained using the Edit Row capability in the Columns grid. It is stored as the Display Width for that column.

That message usually stems from an incompatible version between the MDB file and Access. A run-time version of MS Access does install with QualityStage (and with DataStage, for the documentation tool). It would be unusual that this is out of synch with Reports.mdb but stranger things have happened. ...

You need to increase the size of cache to avoid warnings about "unable to use cache" from being logged. If you're not getting such warnings then your hashed file is able to be fully cached and you don't need to increase the size of the cache. Similarly the size of the row buffering buffer does not n...

That won't work, for at least two reasons. Here's your expression as posted. Can you spot the reasons?

Welcome aboard. :D NULL is always a tricky concept, and therefore should be explicitly handled. If (IsNull(Column1) Or IsNull(Column2) Or IsNull(Column3) Or IsNull(Column4)) Then 'N' Else If (Column1= 'Y' AND Column2= 'Y' And Column3= 'Y' AND Column4= 'Y'&...

Have you searched the forum? You can never be certain from Director whether a job is actually running or not. That its status shows as "Running" means that it has not updated its status to some "finished" status. This can happen, for example, if the process executing the job is killed by an operatin...

Option 1 - do what you're doing in the SQL. This may need slight modification (perhaps CAST function) to make SUBSTR legal as an operand of an date arithmetic operator. Option 2 - also select DB2 CURRENT TIMESTAMP into a third column, whose data type is TimeStamp (or whose data type is VarChar(26) a...

For some folks "performance" (whatever that means) is the primary goal. It's always beneficial to know the most efficient techniques, particularly for large to huge data volumes.

Welcome aboard. :D The five most experienced posters on this site are "premium posters". What that means, in effect, is that their replies are considered to have some value, and a small charge (less than $1 per week) is levied so that you can read the premium posts. This revenue helps towards the co...

There is an error in the Parallel Job Developer's Guide. The NullToValue() function pertains only to the Transformer stage. What you need in a Modify stage is the handle_null() function. You can learn this from the Orchestrate Operators manual - the chapter on the Modify operator.

Fixed-width or delimited format? If it's delimited format, null value can be anything you like. If it's fixed-width format, null value can be anything you like provided it's the correct length (for example "#######" is seven characters long), in which case you don't need to set the null field length...

The default hashed file is getting too big when it approaches 2GB. You can create or resize it in advance to 64-bit if this is likely to be an issue. Trying to minimize what you put into the hashed file (for example no unneeded columns) will help. You can maintain the hashed file as your rows pass t...