DSXchange

Server job doesnt have Dataset stage.
Sequential file will be the only option.
You may need Orchestrate DUMP the dataset into a sequential file either in afterjob subroutine of PX or before job subroutine of the server job.

There is no direct functions available, but you can built your own logic to acheive this. Perhaps you can use buildops for this. First find the total digit, use mod function to get each digit. Create a seperate function which convers the digit into relevent words say 1--> ONE, 2-->TWO... For each di...

sb_akarmarkar wrote:You need to give it condition for insert like InputLink1.NOTFOUND or InputLink2.NOTFOUND and ..... Upto 8 links and for Update NOT(InputLink1.NOTFOUND and InputLink2.NOTFOUND and ... Upto 8 )
above should be fine..

Thanks,
Anupam

Anupam - It is 8 Keys and not 8 links.

adams06 wrote:The column is defined as varchar only.

DSguru2B wrote:Dont change it. Specify that column as varchar. Or you will lose that data.

Why is it complaining for Numeric if it is defined as Varchar?

Sorry I have misunderstood your requirement at first glance. Still you can use Readfirst n row as 1 where you need to process the Header alone. And FirstLine is ColumnName option to skip the Header for processing Detail alone. Optionally you can check the condition in the transformer after reading i...

Is "count(*) as" and the table name alone given as parameter?
If so check whether the parameter is getting passed properly.

rafidwh wrote:And I have 8 key coluns based on that I need to update or insert

In that case InputLink.NOTFOUND should be your simple option, instead of give 8 different condition.

Search for Cartesian product or join.
Create a dummy column say with a value of 1. In both the tables if possible. And make a join betweent the two table on the dummy column during run time. Or do a lookup on the column. Which will force the cartecial product.

Insert the data in the table where the primary key (Parent table) first and later try this job.

Passing parameter to the stages may require to split the job.

Passing parameter to the stages may require to split the job.

Or pass the parameter of the filename from the Jobsequence (if called) with the mentioned method.

To avoid the header mention FirstLine is ColumnName. To get only the header, Give Readfirst n row as 1. Since the detail record has a set, pre process it. grep the line number where "************ DETAIL RECORD BEGINS HERE" presents, and have a head -n to get those detail and header alone. And again ...

Arnd - May I know what is the intention of RESIZE DS_JOBS * * * Will it give out the integrity of the indecies. It seems it requires an Exclusive access to DS_JOBS, so I could not test it now. Kumar REINDEX requires exclusive access as well otherwise the indexes get deleted and not rebuilt. Yes, ev...

Have max(date) and Min(date) from the source, use transformer and stage variable to check till the limit of Max(date) and increment the current date by 1 for each record.