DSXchange

That is a good question indeed. I can't think of a one-liner to solve the problem, though. I would write a short BASIC routine that would do it in a loop, first separate the string into 10 fields delimited by commas, then loop through the results and if the COUNT of double-quotes is uneven then merg...

Your transform stage "EXTRACT_I020_SP_HDW_xfm_sp_data" is having trouble running. Can you try to do a force-compile of your job and see if that either fixes the issue or gives you an error message?

Both Ray and I have been trying to get you to remove the database from the equation - so far you don't know if the SCD is causing the apparent slowness or the databases. Thus you need to make a test job which uses only datasets as source and target (write another job to read your DB selected records...

The OPENSEQ and READSEQ/WRITESEQ don't perform any conversion or interpretation of the data. If you wish to do NLS conversions you will have to explicitly program then. You can check up on the ICONV/OCONV conversion code "NLS{mapname}" to go from one representation to another.

Please post your error message and confirm that you are writing to a dataset and not reading from one.

If all your source and targets are flat files/datasets and not databases then your throughput, even on an underpowered machine, should be in the thousands of rows per second. - how "wide" are your rows, i.e. how many bytes per row? - Is this running in a 1-node configuration and how heavil...

No worries about English - most of the people here aren't native english speakers (including me), but using SMS-Style words such as "u" "plz" "r" isn't appropriate on a professional and technically oriented site.

Phil has it - I wasn't thinking before I responded! I never use lookup filesets.

What sort of a problem are you having going from a Lookup Fileset to a join or funnel stage?

blazDelParis, Welcome to DSXChange. The language used in job sequences is indeed different from that in parallel jobs. You can compute today's dates using the expression CONVERT('-','',OCONV(@DATE,'D4-YMD[4,2,2]')) and that of 7 days ago using CONVERT('-','',OCONV(@DATE-7,'D4-YMD[4,2,2]')) p.s. Plea...

So your actual problem is that your Oracle tables contain valid dates but when you read those values into DataStage you are getting NULL values? That means we can ignore Sybase completely for the moment. What Oracle stage are you using? If you do a "view data" do you get null values for da...

Absolute numbers are not helpful since they depend upon a lot of factors.
Can you change your sources to flat files or data sets and do the same with your output and then measure the time. That would narrow down the possible problem sources appreciably.

The definition and calling of routines is totally different between Server and Parallel jobs. You posted in the Server forum but marked it as parallel. Which is it?

behrouz - No poster is being paid to help here, it is all done on a volunteer basis. Running a website costs money and the membership fees help defer costs. While you seem to feel that all help should be free, I note that you have chosen not to do so - had you posted a full response it would have be...

You should stop and re-start DataStage in order for changes to the dsenv to affect all processes; but the question would remain as to why it worked before (unless the entry was changed).