DSXchange

Do a man on remsh

Remember, a hash file is a presized parking lot where the stalls or spots are big enough for a couple of cars each. With a minimum modulus the hash file will always be a certain size when empty. You don't know when it's filling up because the data is being distributed within that parking lot. This s...

I wanted to share this private email with everyone so that we all benefit from the discussion: Just noted your response in the DSXchange to request for a method to transport files via method other than FTP. You say never to use the FTP stage as it is a gimmick. We use this stage fairly extensively t...

Never use this stage, it's a sales gimmick. Use an FTP command line script to move files locally, and then distribute when done. The FTP stage is not practical because the only reason to use it for large files is to prevent copying the file locally, except, a large file will take FOREVER to read usi...

Sure: 1. Have you worked with DataStage? 2. Have you worked with any ETL tools? 3. Have you worked on a data integration project? 4. What is your experience with data warehousing? 5. What is the largest table you have populated? 6. What are the key factors in developing code in a managed environment...

Unfortunately, you have to use a common data type model, and that is what DataStage is doing. When importing definitions, they are translated into the common model. When loading data, the metadata is cast into the target table type. Whatever DataStage imports as metadata is what should be (in most c...

DS runs under the engines environment, not the user running the job. Testing a script from the command line is comparing apples to toaster ovens. You must look at the environment your jobs execute under and make sure all variables and paths are appropriate.

Sure, for the delimiter use 009 in the stage definition, which DS will interpret as ASCII(009).

Yes, because it keeps track in a local hash file what the last value assigned was. However, this is unbelievably dangerous in approach and design. If the hash file was corrupted or reset, or, you manually inserted/deleted/fixed data in your table, you run the very real risk of screwing up your data....

You can't read a file that fails to open.

Here's what I would have written. This is untested, but you should get the idea: GoodCount = 0 BadCount = 0 FileList = TRIM(InputArg) CONVERT "," TO @AM IN FileList FileCount = DCOUNT(FileList,@AM) For FilePtr = 1 to FileCount FileName = TRIM(FileList<FilePtr>) Unix...

Are you needing to truncate the table and then insert rows? Or, do you need to update existing rows with a whole row replacement? I don't understand the need to delete a row you could whole row update.

Here's a solution that doesn't have the performance and maintenance issues (secondary indexes) of a UV/ODBC stage:

http://www.dsxchange.com/viewtopic.php?t=84998

Don't EVER do what you did again. Next time, simply use DS Administrator to remove a project. The simplest solution, since you've corrupted DataStage, is to wipe your installation and re-install the engine and all projects.

If you mean zero rows returned, then use job control API calls and the check the link statistics after the job has run. If no rows were processed, then take appropriate action.