DSXchange

Hi all,

Pl help me out, i have a DS job, which writes in to the tables and sequential files based on the condition.
But it is writing into the tables correctly, but it is not writing into the sequential files.
I want to know, is there any row limit in DS, as the size of flat file may goes max of 1677 characters. Average of 800 characters

If i see the logs in Director it is showing few records loaded, but it is not writng anything in to the file

if i replace with the dataset, it works fine.

thanks in advance
basi

No limits of the type you are looking for exist. How are you ascertaining that the records aren't being written to the sequential file? Does the designer "view data" show anything? Does the Job Director log actually state the n-rows were written?

ArndW wrote:No limits of the type you are looking for exist. How are you ascertaining that the records aren't being written to the sequential file? Does the designer "view data" show anything? Does the Job Director log actually state the n-rows were written?

Director log shows 3 records loaded into this stage, But if i see from the
Designer View Data i am getting the following error
" Import complete; 0 records imported successfully, 0 rejected."
Even i tried to see from the Unix, it is showing 0 bytes size of file

Pl help me out

Basi

Change the stage to a peek stage and see if the data looks like what you expect; perhaps some values are incompatible with the PX settings (nullability is a common cause) and are keeping the records from being written. Usually these messages can be seen in the director (but sometimes they are very well hidden among the other messages).

ArndW wrote:Change the stage to a peek stage and see if the data looks like what you expect; perhaps some values are incompatible with the PX settings (nullability is a common cause) and are keeping the records from being written. Usually these messages can be seen in the director (but sometimes they are very well hidden among the other messages).

If nullability is the constriants dataset will also restrice the row or atleast warning would have poped up.
But as Arnd said, peek would defenitly help.

regards
kumar

hi,

The reason is you are writing nulls to a sequential file without handling nulls. That is the reason eventhough the log says 3 records are loaded but you are not able to see any record in the view data nor in the unix side.

To handle this, go to the sequential file, format tab, field defaults. select the property NULL field value and give any random value such as "@" and run the job and then try viewing the data. now where ever nulls are present the symbol would replace and you would be able to view the data both in unix and in datastage. In datastage the fields will be shown as Null. in unix it will show as "@"

Madhavan VM wrote:hi,
To handle this, go to the sequential file, format tab, field defaults. select the property NULL field value and give any random value such as "@" and run the job and then try viewing the data. now where ever nulls are present the symbol would replace and you would be able to view the data both in unix and in datastage. In datastage the fields will be shown as Null. in unix it will show as "@"

setNull() works too. its alwasy better to trun it back it to what it really is