DSXchange

Create a text file in the form of "search value | return value". Write a function that takes two arguments that reads the text file and loads into a COMMON on first invocation. Read the file in as a two row array. In the first row, put the values to search, in the second row, put the value to return...

You gain performance because of a massively parallel distributed processing model. You gain processing performance because you link multiple servers together into a "grid" like processing model. For a single Windoze server with limited cpu count, you'll probably not gain the distributed processing i...

I think you didn't read my recent response carefully. as the Aggregator wants DATE in the internal format, which is stupid. I was commenting on the Aggregator stage requirement to process DATE values in the internal DS BASIC format. As for this thread, your attention wandered, you posted one issue, ...

Run all data thru an Aggregator and use LAST derivation for all columns with no GROUP set.

You have not posted an error message. First we fix your SQL, but you switched to using a different method while we are wasting time on your SQL. Since you have not posted the error message and we can't read your mind, I'm guessing you're getting a data type error. If you use the Aggregator stage wit...

Okay, forget the SQL file suggestion, I didn't post my reply and missed the 0 rows output with array size 70.

You've got a bug.

Add a Sequential file stage to output all rows from the Transformer to put to bed any issues. I've never seen this issue with array size when querying. If the Seq file doesn't have all of the rows, it's a bug. I've used Array Size in the past as a tuning method and never saw dropped rows.

Your SQL doesn't work. You either need to add more columns to the GROUP clause or use a subselect. You also need to do the MIN on the data before converting it to char. You also have a trailing comma in the column list that won't let the SQL work. Either: SELECT PERIOD.CPY_0,PERIOD.FIYNUM_0,PERIOD.P...

The question should be: do I use MV's to maintain aggregate (summary) tables and collapsed dimensions or write ETL? The answer is: depends on your volume of data, processing window, hardware available, skill of your team, and time available to develop. MV's are easy to use, but problematic. There ar...

Make sure your Sequential stage in the OCI-->SEQ job is setup to honor quotes if your delimiter is in the data. If not, don't use quotes in the setup. If so, you then need to make sure your data does contain quotes that will foul up the detection of column separation. In addition, data containing lo...

Please use the Search facility. This has been covered extensively for years. Here's a post you might find interesting:
http://www.dsxchange.com/viewtopic.php?t=85578

A job is responsible for updating its own internal status, config, and log file during execution. If the job dies tragically, those files remain in their last state. Clearing the status is ONLY used to cleanup a job that didn't die gracefully. You can clear the status file while a job is running, wh...

It's the total time to get the LAST ROW that matters. If you do SELECT * from a table, it is no indication that Toad can extract 100M rows in 1 second, it just means that's how long it takes to find the first row to output. In DataStage, you're seeing the time it takes to find the rows and send them...

DSguru2B has a good solution for you to use.

Are you writing the mapping documents? Where are your business analysis and subject matter experts? They're going to have to describe the data flow and transformation rules. As for order of population, "outer-in" is the method to describe working from the tables on the sides of a datamodel to the ta...