DSXchange

Welcome Aboard.
You need to add -job status somewhere in your dsjob command before giving the Project name. That will return you the true job status.
The possible return codes can be found in the Basic guide (.pdf) or clickhere.

You have to go about storing it in a file or a dataset. Server jobs have UserStatus areas. PX jobs dont.

That is why i pointed out Vincent's FAQ entry about generation of keys in a px environment. It takes into consideration the number of partitions and current partition number. Refer to my previous post, it has reference to a link.

Like always, i missed that part in the original post, i have a problem

The following doesnt work for you Ans = "" cmd="SELECT JOBNO FROM DS_JOBS WHERE NAME NOT LIKE '\\%' SUPPRESS COLUMN HEADING COUNT.SUP;" Call DSExecute("UV", cmd, JobNoList, ReturnCode) If ReturnCode<0 Then Ans=@NULL End Else Ans = Convert(@FM, ",",...

No stage variables. Again use the concatenation method.

If your Col4 is varchar, nothing more is needed, if its integer then either use AsInteger or StringToInteger() to convert the concatenated string to integer.

It really depends. On a higher level, Error Table will have the data that created the error and Log Table should be details on why it got rejected and other time information like you said. But it really depends upon whats the "in-house standard".

Handle situations where the input can be null. Thats what it is complaining.

In.Col[1,2]:"-":In.Col[3,2]:"-":In.Col[5,2]:"-":In.Col[7,2]:"-":In.Col[9,2]:"-":In.Col[11,2]:"-":In.Col[13,2] aka Substring Concatenation method.

Does it remain constant. Are you always going to get 14 characters or it could be less. If yes then use substring and concatenate.

In your above routine use this and get rid of everything after your DSExecute() command.

Welcome Aboard
What is the max number of characters? Do they remain constant. If yes then you can simple use substring() and concatention technique.

DataStage Help.

Sure. Read a record from your source file, get the first column, find all records that match that col1 in the second file, lets call it dbfile. Now take one row at a time and do check each and every column, if all columns match well and good, write it to your output file, if not then check if the fa...

No worries Whale, I fell for Ereplace() too. We are all here to learn
Right Ray ?