Oracle Enterprise read method

[somu_june]

Hi,

Iam using oracle Enterprise stage to read data from a table


SELECT * FROM EMP WHERE EXISTS
(SELECT DEPT FROM DEPT A
WHERE A.DEPT_NO = EMP.DEPT_NO)
ORDER BY EMP.ENAME


When I am using above query to run my job to extract from oracle enterprise stage it is extracting only 5000000 (half million) instead of 1 million records . When I ran the same query in SQL* PLUS it is giving me 1 million.

I tried the option of ARRAYSIZE in DB OPTION and changed to 4000 but no help. I ran the same job for less volume of data like 400000 the job is working fine ( with out changing any seetings like array size) but when I ran huge volume of data it is not extracting all records from the stage.

Is there any setting that we need to handle like ENV variable.



Thanks,
Somaraju.

[vinothkumar]

How are you determining that your stage retrieves only half million records ?

[somu_june]

Hi,


From Job Monitor and Job log. I determined the no of records Iam extracting




Thanks,
Somaraju.

[ray.wurlod]

Did you look at one node of two, perhaps?

[somu_june]

Hi Ray,


Iam reading it sequentail so it will have only one node





Thanks,
Somaraju

[bollinenik]

Hi,
Try find how many number of records are there in the file from unix box with unix commands or in some other way, insted of depend only on DS link counts.

unix: wc -l <filename>

[somu_june]

Hi,

When I ran the same job from Development (Datastage) server pointing to Preprod Oracle server, job is working fine and when I ran the same job from Preprod (Datastage) server pointing to Preprod Oracle server job Iam facing the above problem as I mentioned and I checked all the environment variables from Datastage Side and every thing looks fine.

Is there some thing I need to look from unix (sun solaris) side.

Thanks,
Somaraju.