DSXchange

Hey i got some problem

I have 76 records in database and when i want to dump that into staging

it is populating only 68 records.

I am missing 8 records

I have no idea why these records were omitted by DS

will there be any record level locks which avoids DS to access the records

help me out

Thanks

How many records does your job log show coming out of your source database- 76 or 68? What is the key on your staging table? Could you be overwriting duplicate records on staging?

hey Arnd,

my job log shows 68 records. and the thing is there are no duplicate records to overwrite

and the key is the ROW_ID

I AM NOT ABLE TO UNDERSTAND WHY THIS IS HAPPENING

thanks for the reply

This is NOT the way to explain your problem. Please be elaborate when you talk about your problems.

In your case, you have NOT given the type of database, the stage that you are using to extract these records.

Any constraints used while extracting. Please always give a better description of your jobs so that you get a good solution.

If you give only tiny bits of information. You will be at the losing end.

I'm saying this for your own good.

Thanks,
Naveen.

What is your target? For example if you're loading a hashed file, duplicate key values are destructively overwritten. Do you have a constraint on your output link in the job design? Without knowing these kinds of things, it's really difficult to diagnose your "problem".

Naveen thanks for telling it

my source is siebel and target is oracle

i am doing dump but i am not able to dump entire records from siebel
i have 76 records in db but ds is picking only 68 records and dump to oracle.
As far as my knowledge there are no duplicates

and the constraint i used id update_date>1/1/1900 and flag='y'


helpwill be appreciated

No Problem. Glad that you understood my message.

So, you said that the constraint you placed was:

update_date > 1/1/1900 and Flag 'y'

Does your siebel source have 76 records with the above conditions met?

I'm assuming that only 68 records satisfy this constraint.

Please look into it and let us know.

Thanks,
Naveen.

Have you tried dumping it into a flat file to see if you get all of the records in the file?

This would tell you whether the records were rejected by the data base or not.

Hi,

Lately, I have also this kind of problem. I had to transfer records from DB2 to SQL Server. Some records were just ignored. When I look in Director, I saw a lot of warning. It seems to be a language mapping problem. Did you look at the log in Datastage director? What does it say? What kind of data do you try to ETL? Is there any special character in it? If it is so, do you NLS enbaled?

Good luck

What happens when you run your SQL using your favourite SQL Utility?

Cut and paste the SQL from your ODBC/OCI query stage into the SQL utility (what backend is Siebel hosted on? Oracle?). Do you get 76 records or 68? If you get 68, the problem is with your SQL query. If you get 76 rows, the problem may be somewhere in your DS job.

Open the job in Designer. Turn on 'Show Performance Statistics' by right clicking on the canvas. Are you seeing 76 rows coming from your source or 68? If you see 76 rows, you are losing them somewhere in your job. Observe which stage has 76 rows coming in and 68 going out. That's where your problem is.

HTH
Rob Wierdsma

Just another debugging tip, try hanging a reject link off your oracle enterprice stage and set the reject flag to true inthe stage.

This captures any rows that were rejected by the database for some reason.

The best way is to see truely what your source query returns, ensure there are no constraints that are dropping due to conditions not being met and review the director logs to get to the issue....


Arvind