DSXchange

Check the value properly,whether the value is null or empty.

You can use a transformer stage variable also to split the records

You can use a modify stage with the following function

int64_from_decimal

You can also use transformer stage in both the source to derive a user defined col and then do a look up on the basis of those columns

Depending upon your data volume use join stage or look up stage and then a filter stage

If only partition related issues are handled in the transformer then actually no need to implement transformer.This can be handled in any other stage.
Ex-For remove duplicate stage use a sort before it with hash partition and then make the partition in the rdup stage as same.

1. In a transformer or modifier stage define a stage variable say sVar1 seperate out the day and month part from the current date.Now concatinate it with current Year -4 on proper date format.This value is converted into date and stored into sVar1. 2. Now compare between sVar1 and current date with ...

One thing may works.Disable the RCP

Check the nullability defined in the Output column.It should be null.

First use a sort stage ,sort on group keys.Make the partition to hash,with the same order as the sorting order.

Now use a aggregator stage with partition as same.Hope it will works.

I can give you an idea.Don't know whether it will help or not. 1. Just write a shell script to write with 1,2,........,N in a col say COL1 in a text file 2. Again add a col say COL2 with some default value say 1 using the script. 3. Now call this script in your sequence job. 4. Then in your parallel...

1. Your Propagation flag need to clear all the time
2.For look up stage it is absolutely fine.Because in entire the reference is being copied to all the node which is necessary

1. First need a dummy col with any value say 1 you can do it in a transformer stage or col generator stage 2. Need a aggregator stage where you determine total salary of the employee 2. Now use a sort stage and sort the salary in descending order and then remove duplicate on the basis of dummy col Y...

It is better to use hash partition whenever you are working on group of data like aggregator,sort,remove duplicate.

If the no of fields in the source file get changed then the whole job flow will be affected.in this case what is the exact scenario?