DSXchange

The easiest solution call up the command before starting your current run so you have the start number of the previous one. This would be the easiest solution. The other ones all involve writing some code either in your shell to use dsjob -logdetail and looping through all the entries.

I'll be the second to second that

Try using the actual IP address as a test instead of the hostname. You cannot use the "omit" option with a UNIX login, you will always need to supply a userid and password.

An "API" is an Application Programming Interface and technically all of the stages listed are API's into DB/2 and TeraData respectively. I think the common usage for the API would be the DB2/UD Enterprise stage and TeraData Enterprise stage.

There are various ways that you can do this. How you solve it depends upon exactly where you calculate this date and where you call up the external command. The easiest solution might be to write this date value to a file (one row, one column) and then you use the contents of this file indirectly in...

Yes, the DataStage engine should always be brought up with the system boot.

Just one partition/node with that configuration

add a transform stage and add a column with the derivation of "LEN(In.OracleColumn)" and it should be 1 and not 3

How is this job taking a "backup"? Perhaps this job is waiting for some resource to be released and therefore looks like it is hanging. Also, how did you kill the job?

What is the column defined as in Oracle (not in the oracle stage, but in the describe of the database itself)? It looks like a CHAR(3) definition and it is padding with CHAR(000); in this case DS will use the actual data and not the metadata defined in the column.

a=str1("abcdefgh") The function "str1" is not a builtin DS function, so I don't know what it is supposed to do. substring(a,3,2):@vm:@sm will take 2 characters from string a and append a @VM and @SM {value mark and subvalue mark, CHAR(253) and CHAR(252)}. These two "specia...

I you do multiple job compile and select ONLY the one sequence, does it also fail? If yes, then you need to report this to your support provider {and you have a simple test case to submit to them}. Does this sequence fail to compile every time or is it a sporadic error?

I should have also added that you can look at Kim's etl utilities, which also collect some informatinon. At present I am doing something similar for a large implementation. This involves quite a lot of coding to make sure that no logging information is lost and it collects logging information for 8 ...

If you need to extract this information immediately after every run you would have to either: a) use an after-job subroutine that is fired off regardless of job status b) If you use an external scheduler, create your own "interlude" dsjob call which not only starts the job but also starts job loggin...

The mark characters are just non-displayable ASCII representations that sometimes have special meanings (depending upon where they are used) in DataStage. @FM are used in dynamic arrays (another way of saying strings in DS) to separate fields. They need to be carefully used for hashed files, if you ...