DSXchange

Hello I have downloaded the datastage tutorial (ver 3.6 !) from http://www.oninit.com/manual/informix/english/docs/36dstage/6471.pdf I do not however have access to the source files and setup scripts. If anyone could provide me with these, I would be grateful. I seem to remember that these files wer...

De nada, I just noticed Ray's got more points than me.

Define your primary keys in the output columns of your filter stage.

Use an Execute Command stage in your sequencer. Issue the command "SLEEP 300"

Ok, apart from the = issue in Value6, you have also assigned values in your parameters from 3 to 8 e.g. your code reads Value2=Field((File<3>),'=',2) surely you mean this to be Value2=Field((File<2>),'=',2) ... Value7=Field((File<7>),'=',2) An easy fix for Value6 would be Value6=File<6>[8,999] or Va...

Can you paste the value of your input file ?

I tested this code and it works fine. Your file must either be empty or not contain '=' delimiters. Do a call DSLogInfo(File,"File") and examine the contents. Of course if you require "=" to be included in your parameters you will have to use an index command to substring your pa...

Please note a modulus of size 1 is probably not a good idea. The minimum size I'd use is 997.

Have you tried a bulk load? Your flat file could load directly into oracle without the need for datastage at all. Also, remember to check "Allow stage write cache" for your hashed file. If you know the key type of the hashed file you could get a marginal improvement in performance by makin...

What happens when you "view data" ?

Have you defined your meta data as varchar unicode?

I've seen this problem on 8.1 as well. Try reading the data as a char and then do a conversion on it. You will probably be able to see the data as a string with a signed value on its end e.g. } would be a 0. I'd be interested to hear if anyone else is struggling to upgrade CFF to 8.1.

This is actually quite an advanced topic but try to think in terms of a musical orchestra or a choir... Set up the environment variable $APT_DUMP_SCORE in your job. This will display the "score" in your log file. This score could also be considered as the conductor node which shows the ent...

Er, you're trying to put NULL into a sequential file? Good luck.

Read the entire string as a single varchar(1000) in sequntial file mode

Map your fields in a transformer:
ColE : if index(input.col,'COL E=',1) > 1 then field(input.col[index(input.col,'COL E=',1)+7,99],'&',1) else ''