DSXchange

I used to do a 'tricky' conversion with Informix, reading an EBCDIC flat file into the DB as an external table, then writing it back out to a different ASCII flat file. The key to success of course was the metadata, to correctly describe the two external files. Given DB2's roots and parents, I'd be ...

If your CDR's are in ASN.1 format, then the best solution is to use an ASN.1 decoder, implemented in some other language (C++, Perl, Java, etc.). AFAIK, DataStage does not natively support ASN.1. Trying to do one yourself in DataStage will likely not result in the kind of performance that CDR proces...

No, I don't think so: select months_between(to_date('2005-02-01','YYYY-MM-DD'),sysdate) lo, months_between(to_date('2005-02-28','YYYY-MM-DD'),sysdate) hi from dual gives: lo hi ------- ------- -1.466 -.595 We want to process all records in range 2005-02-01 to 2005-02-28, but above would exclude the ...

That's the whole point, it compares the month/year of the candidate record to month/year of previous month, and processes if the same (isn't that what you want?) So if you run it right now: select to_char(add_months(sysdate,-1), 'MMYYYY') from dual You get the month/year of the previous month (year ...

You can use the Oracle add_months() function to get previous month/year: prev_month_year = to_char(add_months(sysdate,-1),'MMYYYY') So your condition would look something like this: WHERE to_char(FIELD_HISTORY.MODIFY_DATE, 'MMYYYY') = to_char(add_months(sysdate,-1), 'MMYYYY') Remember to add checks ...

Why not let the target DB engine do the heavy lifting?
First load all the rows into table A.
Then select unique from A into B, generating your B sequence number as you go according to the capability of your DB (Oracle sequence, etc.)

Carter

Another option if the data set is not huge, is to use the Merge stage. This stage reads flat files, and will perform various Left, Right, and Inner joins to create output datasets. Query each DB, creating a flat file, then use the merge stage to join them. Some people find ithe Merge difficult to se...

Use an Oracle Date formatter like 'MONTH DD YYYY HH:MI AM' with your to_date().

Carter

So, is this a translated COBOL file or what? Is the FD available?
Perhaps you can deal with it as Complex Flat File.

Carter

Since you didn't post *exactly* what you are submitting on the command line, it's harder to answer your question. Remember that the shell gets first crack at anything you submit on the command line. So be sure thet there are no alias definitions or whatnot that match in your invocation. Make sure th...

If you are talking about commercial job schedulers, then most have some notion of resource allocation. You can, for example, define a resource, and it's max load (100 for example), and then associate a resource load value for each job (10 for example). The scheduler will wait until sufficient resour...

Oops, forgot to say that you should document such kludges *very* thoroughly.

Carter

You didn't specify what database. One way I have used with Oracle OCI stage, is to dynamically create the Sql statement into a named flat file, then specify that file as the source of the query in the OCI stage. ex) Derive the date value into stage variable dateSvar. Write the query into flat file D...

Do as Ken suggested, writing the query output into a hash, but reverse the ORDER by clause so the the LAST record written into the hash is the FIRST one that you want. When the query completes, what remains in the hash will be all of the FIRST records.

Carter

At my client site, they uses ClearCase. It doesn't play nicely with a *.dsx file, but is OK with *.xml, hence that's what we export for checkin/checkout.

I also prefer *.dsx for my own uses.

Carter