Page 1 of 2

job aborting in merging of two files

Posted: Fri Jul 01, 2005 12:30 pm
by harithay
Hi Everyone,

i am merging two files (inner join) , with 2 key columns in both files.

i am not doing any tranformation in transformer.

i am just loading to sequential file .

when i execute the job it is giving following error.

Abnormal termination of stage nsar_merge..Transformer_1 detected

i could not solve this one. any suggestions.

Thanks

Posted: Fri Jul 01, 2005 5:07 pm
by ketfos
Hi,
1.Are any records being written in the target ouitput file?
2. What Update Action option you ahve checked in the target file stage?
3. Are you able to view the data uisng the VIew Data button?

Ketfos

Posted: Fri Jul 01, 2005 7:29 pm
by ray.wurlod
What did searching on this forum for "abnormal termination" reveal to you about how and where to look for diagnostic information and/or likely causes of this particular message?

Is it correct to discern from the error message that the job name is nsar_merge and that there is a Transformer stage called Transformer_1 in the job design? Can you post your job design, which I infer is something like

Code: Select all

Merge stage  --->  Transformer stage  --->  Sequential File stage
Does the problem occur in row #1 or later? Can you use the Debugger to step through the job?

Posted: Sat Jul 02, 2005 1:19 am
by Sainath.Srinivasan
Is that the only warning or critical message in the log? Did you reset the job to see any info in the log?

Posted: Sat Jul 02, 2005 8:22 am
by harithay
Hi ,

thanks for ur reply

i am doing in the same way as you mentioned and the problem is occuring in first row (#1).

0 records are going to transformer.


this is the only warning i am getting .


'Abnormal termination of stage nsar_merge..Transformer_1 detected '

thanks .





ray.wurlod wrote:What did searching on this forum for "abnormal termination" reveal to you about how and where to look for diagnostic information and/or likely causes of this particular message?

Is it correct to discern from the error message that the job name is nsar_merge and that there is a Transformer stage called Transformer_1 in the job design? Can you post your job design, which I infer is something like

Code: Select all

Merge stage  --->  Transformer stage  --->  Sequential File stage
Does the problem occur in row #1 or later? Can you use the Debugger to step through the job?

Posted: Sat Jul 02, 2005 8:25 am
by harithay
Hi,
thanks for ur reply sainath,

this is the only meaage i am getting in the warning .

Sainath.Srinivasan wrote:Is that the only warning or critical message in the log? Did you reset the job to see any info in the log?

Posted: Sat Jul 02, 2005 8:43 am
by chulett
But what about his second question? Did you reset (not recompile) the job and get any additional messages in the log labelled 'From previous run...'? More clues there if it shows up.

Posted: Sat Jul 02, 2005 10:29 am
by harithay
Hi,

i did not get any additional information.

if i reset and again run it , i got same message

Abnormal termination of stage nsar_merge..Merge_transformer detected
chulett wrote:But what about his second question? Did you reset (not recompile) the job and get any additional messages in the log labelled 'From previous run...'? More clues there if it shows up.

Posted: Sat Jul 02, 2005 1:50 pm
by Sainath.Srinivasan
The error appears to raise abort on Transformer_1, which I assume is the first (and only) transformer in the job.


Can you check whether there is any logic embedded as part of the transformer.

Posted: Sat Jul 02, 2005 7:27 pm
by ray.wurlod
The error may actually raise on the Merge stage. However, being a passive stage, it has not reported the error but its caller, the active (Transformer) stage, has reported it.

Please post exact details of the join you have attempted including the definitions of key columns in both source files.
Do these files have column headings in the first line? Do these files have null or empty values in any of the key columns?

Posted: Sun Jul 03, 2005 8:21 am
by harithay
Hi ,

Thanks for reply.

i am merging(pure inner join) two files. both files are in same format. the key columns both of them are fund_name, cik .
there are no nulls in the key column (in this case i am dealing with test data first file with 15 records and second file with 24 records ). i need to merge based on key columns(if it matches)
both fiels have column headings
these files don't have any null or empty spaces in key columns.


ray.wurlod wrote:The error may actually raise on the Merge stage. However, being a passive stage, it has not reported the error but its caller, the active (Transformer) stage, has reported it.

Please post exact details of the join you have attempted including the definitions of key columns in both source files.
Do these files have column headings in the first line? Do these files have null or empty values in any of the key columns?
Thanks

Posted: Sun Jul 03, 2005 9:52 am
by elavenil
What kind of transformation done in the transformer stage?

Check any invalid characters in any one of column in both files!.

Regards
Saravanan

Posted: Sun Jul 03, 2005 10:19 am
by roy
Hi,
Most likely your problem will be resolved if you rebuild (not copy paste) your mege stage a new (I've been this way before :().
I've had my share of this behaviour.
Be aware performing merges of huge files might proove to be:
1. slow
2. fail if you don't have enough resorces.

IHTH,

Posted: Sun Jul 03, 2005 1:38 pm
by harithay
Hi ,

Thanks for reply

i did redesign of job (rebuliding merge stage) but same problem.

i am designing like this

merger- ----> transformer- - ----- -> sequentail file.


i am not doing any transformatios in tarnformer ( just merging in the first stage).

even after redesigning (reduilding merge stage) i am getting same error.

Code: Select all

Abnormal termination of stage Merge_Testing..Transformer_1 detected
roy wrote:Hi,
Most likely your problem will be resolved if you rebuild (not copy paste) your mege stage a new (I've been this way before :().
I've had my share of this behaviour.
Be aware performing merges of huge files might proove to be:
1. slow
2. fail if you don't have enough resorces.

IHTH,

Posted: Sun Jul 03, 2005 3:19 pm
by Sainath.Srinivasan
Did you confirm that the data meets the requirement?

Also whether the method works with test data you know is valid.