DSXchange

Hi,

Hope you all had very good Christmas holidays.

I have created a sequential file, which was created using Server job and i need to use the file in PX job using sequential file stage. Field delimeter & quote character has been setup as it was created in the server job but it shows only one record out of 10 record. Noticed that some of the columns having null value in the other records and i suspect this null is causing the problem.

Can you someone share me your ideas & suggestions to view data in seq file stage in PX.

Thanks in advance.

Regards
Saravanan

Saravanan,


if you have only one record out of 10, i suspect you did not specified the "Record Delimiter" or "Final Delimiter" field at the record level, which should be something like 'Unix NEWLINE' or 'end'.

Benouche

Hi,

Thanks for your reply.

I do not think it is not the case. Because it showed only the third record, which has appropriate values for all columns. So Record delimiter could not be the issue and moreover i already tried setting the record delimiter and final delimiter as you mentioned in your email. But the result was the same.

FYI, i could see all the records when all the columns have some values. That is why i am suspecting null value causing this issue.

Pls share your suggestions & ideas.

Regards
Saravanan

Did you check the 'Default' value ? you can specify it by doubleclick on the specific field in the column tab.

elavenil wrote:I have created a sequential file, which was created using Server job and i need to use the file in PX job using sequential file stage. Field delimeter & quote character has been setup as it was created in the server job but it shows only one record out of 10 record. Noticed that some of the columns having null value in the other records and i suspect this null is causing the problem.

Man, I wish I have DataStage here to verify a few things, but you need to remember this:

Despite the efforts done, PX is NOT the same as Server for 6.0. It got a bit better with 7.0, but total integration is not yet there. They may look similar, but the details are devilishly different.

As others have noted, it is vital to indicate record delimiter, and final delimiter. Also, you need to ensure that your input fields can retain the data provided - it needs to be defined with a length equal or larger (and set as varchar if it is a character string.)

If you have some patience, and wait until Monday, I can help you with the specifics that may be affecting your job. Go ahead and leave a private message to remind me.

-T.J.

Once you import the seq file definition in Manager, double click the icon and use the format tab which has a synch to Parallel.
It will show you that some properties do not apply to PX say ok

Then remember that PX can operate with Datasets better and you could use the peek stage to view data from the sequential file.
There will be a show OSH if u have enabled OSH visibility.
PX performs an import opn each time it encounters a seq file!!

Hope this is of help to u.

Regards
Srini

srinivasb wrote:Once you import the seq file definition in Manager, double click the icon and use the format tab which has a synch to Parallel.
It will show you that some properties do not apply to PX say ok

Then remember that PX can operate with Datasets better and you could use the peek stage to view data from the sequential file.
There will be a show OSH if u have enabled OSH visibility.
PX performs an import opn each time it encounters a seq file!!

Hope this is of help to u.

Regards
Srini

Hi Srini,

What do you mean by 'PX performs an import opn each time it encounters a seq file'? Is this the reason why retrieval information from sequential can be so slow?