DSXchange

Author(perfman),

Regarding your problem, there can be for now two solutions with the touch of flexibility by one hand.

1. If the format of the Input data is flexible to be changed as having blankspace, the above mentioned method can help you.

2. If the format of the Input data is NOT flexible to be changed as having blankspace, get the TimeStamp as Varchar and handle null in the Transformer Stage then convert them into TimeStamp.

ray.wurlod, chulett

As you both have been suggesting perfman to create an Environment varibale which he pointed out, I now have an unstablity in understanding the Environment variable.

How the manual creation of an undefined-but-documented Environment variable(starts with APT, for parallel Extender) using DS Administrator would give the result of making the expected effect on Job/Stage/Link Objects as good as the defined Environment variable(starts with APT) in the project?

Consider this case:

Suppose i do not have an Environment Variable APT_BUFFER_FREE_RUN available in the DS Administrator as already defined and so I create an user-defined Enviroment Variable APT_BUFFER_FREE_RUN in the Administrator with Default value 0.75(String) then I use that Environment Variable in my Parallel Job, How the value of this variable would give an effect of executing the whole Job in 75% buffer free run?

Please clarify this.

Default values are coded within the osh executable and used if the environment variable is not set. If the environment variable is set (anywhere) then its value takes precedence over the default value.

But, i feel my question is still alive, as, for me, your words did not seem to be highly related the question.

Maybe i am waiting for chulett to give his opinions.

I would like to redefine my question as simple as below.

Is there anyway to create(if not found already) an Environment variable for specific purpose such as - Buffering, Debugging, Partitioning, Reporting and so on as the same function as parallel Extender defined Enviroment variable?

No. You must use environment variable names as documented in the DataStage or Orchestrate manuals. No others will be recognized (unless in your own expressions - and there you can not affect "buffering, debugging, partitioning, reporting and so on").

Thanks ray.wurlod.

perfman wrote:Just one question before I do that.

Do I just add it in as I would for any "user defined" environment variable?

I suppose It means the enviroment variable "APT_IMPEXP_ALLOW_ZERO_LENGTH_FIXED_NULL"

chulett wrote:Yes

chulett,

Can we now add that Enviroment variable in the project using Administrator.

There are many environment variables that affect these behaviours. I do not propose to canvass them all here. They are to be found in the manuals, in IBM's Advanced DataStage class, or in my Advanced Parallel Job Techniques class.

ag_ram wrote:chulett,

Can we now add that Enviroment variable in the project using Administrator.

Of course. That's how one would add any environment variable to a specific project.

chulett,

shall i be pleaed to remind you to look at all my posts in this thread(questions to you and ray.wurlod) before anything?

ray.wurlod wrote:OK, try setting the Null Field Value property to "". ...

I set the Null Field Value property to "". It worked for all fields and I could import my file.

Welcome aboard. :D

Thanks. :D