Page 1 of 1
Source Data Truncated
Posted: Thu May 10, 2007 4:17 pm
by DeepakCorning
I have a job which reads from a SQL Server and writes to a Oracle Database. Its one to one job with IPCs in it.
When Irun the job all the character fields from teh source are truncated to length 1 (without any errors) and are loaded in the target liek that.
When I say view data I see the same result. There is no substrings , no logic involved in the complete job.
Let me know if anyone has seen such a kidn of behavior before.
Posted: Thu May 10, 2007 4:53 pm
by ray.wurlod
Version 8 is still very new. Bounce it off your support provider, supply a reproducible case, and let us know the outcome.
Have you specified Char with no length in the table definition? Some databases interpret that as Char(1).
Posted: Thu May 10, 2007 4:58 pm
by DeepakCorning
Sorry Guys , Its not version 8 its Version 7.5. I am not sure how it got that 8.
No , length is provided in the source stage. The source stage is a DRS stage.
Posted: Thu May 10, 2007 5:37 pm
by nick.bond
at the same point you write to ODBC write a sequential file too and check the data in there. This will allow you to narrow down where the problem is.
Posted: Thu May 10, 2007 5:57 pm
by us1aslam1us
On the target side have you checked the display setting, sometimes when you are trying to view data that causes this behaviour.
Posted: Thu May 10, 2007 9:35 pm
by DS_SUPPORT
Write the Output to a Sequential File and Check the data, If you get the complete set of data, then your job is working fine.
In this case Check in the Database whether any triggers or constraint is there.
Thanks
Anand
Posted: Fri May 11, 2007 9:47 am
by DeepakCorning
I removed all the stages in the job and directly sourced the data from the source SQL table in to a Seq file.
I get the same result. Every character field is of lenght 1 in the file , so if a NAME = Mike , then in the file NAME = M.
Posted: Fri May 11, 2007 3:12 pm
by pradeepleon
DeepakCorning wrote:I removed all the stages in the job and directly sourced the data from the source SQL table in to a Seq file.
I get the same result. Every character field is of lenght 1 in the file , so if a NAME = Mike , then in the file NAME = M.
Try this,
Go to the source stage(SQL Server),
outputs tab,
columns tab,
Derivationtab and remove all the column derivations if it has any.
Re: Source Data Truncated
Posted: Wed May 16, 2007 12:05 am
by baglasumit21
DeepakCorning wrote:I have a job which reads from a SQL Server and writes to a Oracle Database. Its one to one job with IPCs in it.
When Irun the job all the character fields from teh source are truncated to length 1 (without any errors) and are loaded in the target liek that.
When I say view data I see the same result. There is no substrings , no logic involved in the complete job.
Let me know if anyone has seen such a kidn of behavior before.
Have u checked the job log in director. see if it gives some warning message and if that message is helpful
Posted: Fri May 25, 2007 9:37 am
by roblew
FYI, I had a similar issue connecting to SQL Server 2005, on DSEE 7.5.1A on RHEL. IBM has issued me a patch, which resolved the issue.
Please contact IBM:
Scase:473187
Ecase: 74157
The patch contains updated versions of the following files:
libdsplugin.so, liborpheus.so, dscapiop.so
-rob
good luck! btw, now I have a strangely similar issue when extracting from SAP R/3, using the ABAP Extract stage. Still working with IBM on this one...