Is Basic Transformer recommended to be use in Parallel Job?

shaminmin26 · Post by **shaminmin26** » Tue Nov 17, 2009 3:47 am

Hi All,

Is it recommended to use basic transformer in the parallel environment. My jobs are hitting the following errors:
Abnormal termination of stage JB_LU_HIST(BASIC_Transformer).#3.BASIC_Transformer
detected.

I had tried to switch them to be running on sequential mode, sometimes it will works well, while sometimes it will be hitting the above errors again.

Currently, i am running 4 nodes in parallel environment. Any ways to solve this problems?

Thanks in advance.

ArndW · Post by **ArndW** » Tue Nov 17, 2009 4:16 am

When you do a reset of your job, what does the log entry "from previous run..." show?

ray.wurlod · Post by **ray.wurlod** » Tue Nov 17, 2009 6:18 am

There must be more information than what you provided. Have you reset the job after it aborts and, if so, is there any additional diagnostic information logged?

srinivas.g · Post by **srinivas.g** » Tue Nov 17, 2009 8:11 am

Not recommended.it is giving poor performances

basic transformer invokes basic language.

Please translate server routine in to parallel routine.

One more way is use server shared container.

ArndW · Post by **ArndW** » Tue Nov 17, 2009 11:46 am

Although I would not recommend it, it is not necessarily slow and certain functions such as some ICONV/OCONV conversions don't exist in PX and are quickly solved in a BASIC transform stage. Be that as it may, the problem here is not whether or not to use the stage, but that it is being used incorrectly in this case. The OP still needs to add further information to answer the question.

shaminmin26 · Post by **shaminmin26** » Thu Nov 19, 2009 9:31 pm

Hi,

When i reset the job, the job will still hit the same errors. Normally, i will have three solutions to overcome it :
1) Clean up the resources
2) Create a duplicate of the current job
3) Restart the services.

I am using basic transformer due to i am using server routines to do transformation of data.

I understand that basic transformer are not recommended to use with parallel jobs but in basic transformer, there are more functions available to use.

shaminmin26 · Post by **shaminmin26** » Thu Nov 19, 2009 9:35 pm

Hi,

When i reset the job, the job will still hit the same errors. Normally, i will have three solutions to overcome it :
1) Clean up the resources
2) Create a duplicate of the current job
3) Restart the services.

I am using basic transformer due to i am using server routines to do transformation of data.

Any solutions that i can overcome this error without changing the server routines to parallel routines?

Kryt0n · Post by **Kryt0n** » Thu Nov 19, 2009 9:47 pm

Have you tried a server job? If for no other reason than to make sure the routine is handling correctly. If it does handle correctly, is performance significantly different not to warrant using a server job?

Cut down your input rows, can you repeat the error? If so, try put a few log messages in to try find at what point it falls over.