Input dataset does not have input field

thanush9sep · Post by **thanush9sep** » Fri Apr 17, 2009 4:54 am

I understand this topic has been discussed, however I am not still clear how to solve this issue.

Dataset1
----------
record
( SRC_RECORD_CODE: string[max=2];
SRC_CODE_SEGMENT: string[max=1];
SRC_LIB_SEGMENT: nullable string[max=30];
SRC_CODE_SURSEGMENT: string[max=2];
)

Dataset2
----------
record
( ID_SURSEGMENT: int64;
CODE_SURSEGMENT: string[max=2];
)

Dataset3
----------
record
( SRC_RECORD_CODE: string[max=2];
CODE_SEGMENT: string[max=1];
LIB_SEGMENT: nullable string[max=30];
SRC_CODE_SURSEGMENT: string[max=2];
ID_SURSEGMENT: int64;
)

Dataset4
----------
record
( SRC_RECORD_CODE: string[max=2];
CODE_SEGMENT: string[max=1];
LIB_SEGMENT: nullable string[max=30];
SRC_CODE_SURSEGMENT: string[max=2];
)

1. while I do look up with Dataset1 and Dataset2, I get an output as Dataset3. At the same time I have a reject link also pointing to dataset4 which gets about 15 rejected rows which I am not able to view by using the option "right click and view dataset4" and gives the below error1.

2. By using "Tools --> Data set management" I saw the rejected row and schema (list of all schema is listed above).

3. Please let me know IF RCP hasanything to do with this.

error1:
--------
##E TOPK 000000 12:11:12(020) <_PEEK_IDENT_> Input dataset does not have field: "SRC_CODE_SEGMENT".
##E TOPK 000000 12:11:12(021) <_PEEK_IDENT_> Input dataset does not have field: "SRC_LIB_SEGMENT".
##E TFSR 000019 12:11:12(023) <main_program> Could not check all operators because of previous error(s)
##E TCOS 000029 12:11:12(024) <main_program> Creation of a step finished with status = FAILED.
-------------------------------------------------------------------------
Instead of using the datset4 for reject, if I use a sequential file then I get these types of error2 and the job aborts.

error2:
--------
Sequential_File_131: Error when checking operator: Could not find input field "SRC_CODE_SEGMENT".
Sequential_File_131: When checking operator: When validating export schema: At field "SRC_LIB_SEGMENT": Exporting nullable field without null handling properties
Sequential_File_131: Error when checking operator: Could not find input field "SRC_LIB_SEGMENT".
main_program: Could not check all operators because of previous error(s)
main_program: Creation of a step finished with status = FAILED.

dsuser_cai · Post by **dsuser_cai** » Fri Apr 17, 2009 8:49 am

I think you have a input column that is not defined in the output or out put column that not defined in the input, I would try to check the meta data, or try to check the properties in the look up stage for reject.

thanush9sep · Post by **thanush9sep** » Sun Apr 19, 2009 10:58 pm

Dear Ray

Could you give me any tips regarding this issue that I am facing.
I would really appreciate your help.

One more point to add to this topic

1. In lookup stage Condition not met=continue and Lookup failure=Reject

Kryt0n · Post by **Kryt0n** » Sun Apr 19, 2009 11:24 pm

DataSet 2 does not have the columns SRC_CODE_SEGMENT and SRC_LIB_SEGMENT for which you have defined the join on. Both input streams must have the same column names for the join columns (I'm assuming you are using a merge or join rather than lookup in order to get this error)

thanush9sep · Post by **thanush9sep** » Sun Apr 19, 2009 11:37 pm

hi Kryt0n

I am not using "Join" or "Merge", instead i am using Lookup and I am using Dataset2 as a "reference" that is why it is not having SRC_CODE_SEGMENT and SRC_LIB_SEGMENT these column come from my source dataset "dataset1"

Kryt0n · Post by **Kryt0n** » Mon Apr 20, 2009 12:26 am

From help files:
You cannot set the mapping or edit the column definitions for a reject link. The link uses the column definitions for the primary input link.

Give your dataset4 columns the same name as primary stream

thanush9sep · Post by **thanush9sep** » Mon Apr 20, 2009 6:02 am

Yes I understand these basic steps, however after doing all the things right in this simple lookup job, I am facing this issue.

I am given all the schema for all the dataset (dataset1, dataset2, dataset3 and dataset4).

Kryt0n · Post by **Kryt0n** » Mon Apr 20, 2009 6:20 pm

so DataSet4 is as follows:

Code: Select all

Dataset1 
---------- 
record 
( SRC_RECORD_CODE: string[max=2]; 
SRC_CODE_SEGMENT: string[max=1]; 
SRC_LIB_SEGMENT: nullable string[max=30]; 
SRC_CODE_SURSEGMENT : string[max=2]; 
)

and not

Code: Select all

Dataset4 
---------- 
record 
( SRC_RECORD_CODE: string[max=2]; 
CODE_SEGMENT: string[max=1]; 
LIB_SEGMENT: nullable string[max=30]; 
SRC_CODE_SURSEGMENT: string[max=2]; 
)

as specified in your original post?

thanush9sep · Post by **thanush9sep** » Tue Apr 21, 2009 9:46 pm

Hi Kryt0n

Thanks now I understand the reason for error.

however for DATASET4 I did not type the column name and datatype, I just connected the reject link to the DATASET4.

Do you have any idea why it is not showing the column name and datatype which should be similar to DATASET1.

I dont know how to change the column name and datatype of a reject link

Kryt0n · Post by **Kryt0n** » Tue Apr 21, 2009 10:19 pm

How I have no idea but that appears to be what it is complaining about... try disconnecting the reject stream from the lookup stage and reconnecting...

thanush9sep · Post by **thanush9sep** » Wed Apr 22, 2009 12:36 am

I did that also "disconnecting and reconnecting"
and then tired by "renaming the job and the dataset"
and then started a new parallel job.
as I had mentioned earlier if I use dataset4 is reject, the dataset4 is created but I can view the data, if I use sequential file instead of dataset4 then the job aborts,

thanush9sep · Post by **thanush9sep** » Mon May 04, 2009 3:12 am

I haven't been able to get a solution for this.

However I changed the name of column for the source similar to that of columns from reference. That is

Dataset1:(Acts as source)
----------
record
( RECORD_CODE: string[max=2];
CODE_SEGMENT: string[max=1];
LIB_SEGMENT: nullable string[max=30];
CODE_SURSEGMENT : string[max=2];
)

Dataset2(lookup dataset)
----------
record
( ID_SURSEGMENT: int64;
CODE_SURSEGMENT: string[max=2];
)

Dataset3(Target dataset)
----------
record
( RECORD_CODE: string[max=2];
CODE_SEGMENT: string[max=1];
LIB_SEGMENT: nullable string[max=30];
CODE_SURSEGMENT: string[max=2];
ID_SURSEGMENT: int64;
)

Dataset4:(reject dataset)
----------
record
( RECORD_CODE: string[max=2];
CODE_SEGMENT: string[max=1];
LIB_SEGMENT: nullable string[max=30];
CODE_SURSEGMENT : string[max=2];
)

Thank you.........

ray.wurlod · Post by **ray.wurlod** » Mon May 04, 2009 4:20 pm

thanush9sep wrote:Dear Ray

Could you ...

Why me? And why when I have not yet responded to this post?

As you note, DataStage expects the column names to be the same for this particular check (same thing - same name). It seems to me that that's a reasonable requirement.

Name changing is free with a Copy stage.

chulett · Post by **chulett** » Mon May 04, 2009 4:24 pm

Why you? Because you are The Man!