DSXchange

The stage variables cannot handles nulls and '' is an invalid date. use a NulltoValue function and define the value with a default date for processing it. When sending the stage variable's value to the output column replace the default value with null

From the logic what i understood is that the value of svx will be added by 1 for every input row. This logic will only work if the config file has only one node. For multiple nodes the stage variable has multiple values on each node and it will add 1 to the current value depending on the partition.

Are you using any stage variable that has the column mentioned. Stage variables cannot handle null. I would say use NullToEmpty or NullToValue functions for handling nulls.

At the end of sequence , Run a job through a job activity . In the job load the table through a database stage and pass the data and time values through a transformer (before the database stage) by calling the appropriate functions.

Well, all of us participating in the forum want to be/are already professionals. There is no harm to set some standards while communicating.

debug the job by adding a peek from the Transformer. The last coulmn should use the same function used to load database and view the data in director. If you are seeing the same data as in database then your function is the problem. (Datastage adds Leading and Trailing zeroes to the Decimal field as...

ashishm wrote:I dont want to do a Join as I may loose records using the JOin Stage.

Please suggest a better way to do it.

Why do you lose records by using a Join Stage? Please Explain.

If you only partition on a key and do not sort, it is 'grouped'. If you do partition and sort on the key then it is 'sorted'. If the sort keys in the sort stage are same as the sort keys in join stage then you need to use - Create Cluster Key Change Column (Don't sort previously sorted) , If the key...

For most of the stages that have a default parallel operator it should maintain sort unless modified. However , the data loses sort when it encounters stages like aggregator.

The data also loses sort when using a sequential opertor is run between two parallel operators.

Did you check the config fie in 8.5 and 8.0.1? I think the number of nodes might be different. This will cause inconsistent results as you mentioned. The hash partitioning you added is the satndard procedure during development, e.s.p for stages like Sort, Join, Remove Duplicate etc.. In this case ca...

DateFromDaysSince() for dates. TimestampFromSecondsSince() for timestamp.

I have described the CLOBs as 'Unknown' datatype. In this case the default datatype assigned is Varchar or Longvarchar depending on the CLOB size. I also observed a unicode extension and it worked fine in my Job.

It is intresting to see people demanding answers as if this is a tutorial. The forum is to suggest and discuss datastage issues. The reply from pandeesh is apt. It would be nice to learn some courtesy.

Do you mean 5 nodes ? link Sort the data in hash partion on the key. ascending or descending depends on the preference.

I am pretty sure it cannot be null. It must be some unprintable character, the best way as suggested by Craig is determine the Hex value and get the equivalent ASCII value and use Convert Function to replace it.