DSXchange

i am just trying to figure out the scenerio where an insert by other process into the same table might be possible during the time between the look up(to find if the key already exists) and Insert (if key doesn't exist). Because we might be running more than one instances of this program on similar...

mahi_mahi wrote:But my concern is it is taking too much time for the SP to finish. Its taking more than 1 and half hour.

The problem as you describe it is not in DataStage. You need to find ways to make the SP go faster. For example, can the updates be coded to operate in parallel rather than consecutively?

PHANTOM is the DataStage command to create a background process, and SQUAWK (or REPORTING) is a keyword that notifies it to advise its parent when done - the report goes into the DSD.RUN file in &PH& folder in the project. So don't worry about these - they are the startup commands for your D...

Umm... it's the Terminator that's new; the Exception Handler has been there since job sequences were introduced. Once the Exception Handler fires you can take an output from it to a routine activity that logs an appropriate message, to a notification activity that sends email, anything else you plea...

A job sequence is merely a special case of a server job, that contains job control code. So is a batch. So is a server job that contains user-written job control code. So you can start these with a perfectly ordinary dsjob command in a shell script. Search the forum for good, robust examples of such...

Without ORDER BY the database server can deliver the first row as soon as it finds it. With ORDER BY the database server can not deliver any row until it has processed all rows to determine which is the first in sorted order. DataStage's clock is running the whole time that the sorting is being perf...

Search the forum for a command called SETFILE for creating the VOC pointer.

Chuck Smith (search again) also has a utility available.

It's a "Henry Ford" choice in the first release - "any database" means one of DB2, Oracle or Microsoft SQL Server.

Historical Note
You could buy a Model T Ford in any color you like, so long as it was black.

One can.

One can even probably achieve it in a generated query.

I am trying to guide the poster in that direction but, for whatever reason, there is some desire there to reinvent the wheel.

Which may lead, instead, to a triangular-shaped rock on a square axle.

No, unless you're prepared to write a pair of Custom stages to manage the memory. DataStage has nothing out of the box. Using a Sequential File stage to write to a named pipe in one job and one to read from that named pipe in the other job is a possibility, but co-ordination of the jobs is crucial. ...

For completeness, in DataStage you could use TimestampFromDateTime(current_date(),current_time()) to generate the timestamp. Use this expression when initializing a stage variable and do not modify the stage variable row by row, so that it is only calculated once. In this way...

^@ is a representation of 0x00 (the ASCII NUL character). You could trim these, you could set APT_PAD_CHARACTER to space (" "), you could pre-process with sed or awk as DSGuru2B suggests.

I suspect you are going to need two essentially identical processes that handle 1 and 2 keys, respectively. As far as I am aware you can not use a job parameter to supply metadata, such as the name of a column or partition key. So you may need to process with a "fake" key - one that's the same for a...

Theoretically if the Unisys ODBC drivers can be correctly configured on your AIX system you ought not need to use any other tools. Edit the .odbc.ini file so that the data source name loads the appropriate driver, ensure that any required environment variables have been set in dsenv, and that should...

Don't forget that you will need read and execute permission.