DSXchange

I'm watching the rain slowly metamorph into sleet here... So even though I can't make it, could you have a picture taken, sign it and then e-mail it to me? I promise not to flog it on eBay and might just use it as my windows background image :D

nexus2me - you have no idea how bothersome it is to see the EXACT same question posted almost weekly. Even though it would seem to be common sense to use the Search before posting a question like this it seems that many, including yourself in this case, do not do so. The NLS_LANG environment variabl...

Martin - GREAT suggestion, that will work. The OP didn't mention if they are willing to use server but if so, using a type 1 or 19 hashed file and updating/appending records will work like a charm. The only downside is performance, but the coding will take only minutes.

I think you get a Windows error message when you do that, something to the effect that the data size exceeds the allocated space and it tells you the number of bytes. There is no solution but to reduce the query size. You can shorten the query (Use x.KeyCol instead of FullyQualfiedTablePathAndNameAs...

Normally I would have just stated "Hire competent DataStage consulting", but several days ago I heard that IBM/Ascential has a software tool that assists Informatica to DataStage (unsure of whether Server or PX/EE) conversions; this tool is not, as far as I know, for sale or use outside of IBM.

There is no easy, builtin way to do this with DataStage. There are ways to code this that are more or less complicated. If you have a known and limited number of output files then you can hard-code the paths in a DataStage job. If your list of customers/output files changes over time or is large the...

I like to FTP the file using a script and using a suffix such as ".in_process" and then, once the FTP is successful, rename the file to the correct name.

did you check to see if the key you are writing is null or if the file system is full?

Try rewriting a test job from your original that outputs the KEY and the SeqFile.single_column[1285,10] as separate columns into a text file and find out why the lookup fails.

Which ones won't change specifically - I haven't seen that and rename my links frequently.

DataStage server jobs do not have "message handling" but the Server portion of the product is designed so that all warnings can be eliminated through good development.

Yes, DataStage PX/EE does this by having the designer put a "reject" link out of the passive output stage. For databases this will contain the SQL error code but in neither case does it explicitly tell you which column caused the reject.

I think I answered your question, but not your problem. It would seem that the column does not contain what you think. Is it a string or do you use OCONV() somewhere? If you were to post the problem and result you will get an answer, otherwise nobody here will be able to guess what is happening with...

There should be more to the error message (something with segmentation fault). Can you post that? Also, if you reset the job do you get an entry in your log with "from previous run"?