DSXchange

Hi All,

We are just trying to understand how to set the limit for the Unhandled pattern which i do not think it is feasible:-)

I have got answers from forum ,to do token/word investigate,group by the pattern jobs on how to identify these patterns, after STAN to have an investigate stage again to find the patterns missed.

Say we have received sample file from src system1 and investigated/identified the unhandled pattern and tried to handle the same.

Now ,there is another sample file and we try to do the whole process.

Do we need to get all kind of sample files from each source systems,to come up with the list of unhandled patterns and then try to customize the rules.

But still when the code is prompted to higher environment, and new pattern is identified then it has to come to development again. is it how it is handled in other projects?

In real-time, is it then we cant' freeze that there will be no patterns that is left unhandled

Can someone put some thoughts.

Thanks.

Every company has its own policies.

In production you run it with the best set of rules you were able to come up with after doing analysis of all the input systems that feed the job.

Patterns that STILL fail are rejected and handled according to your policy, which depends on the # of records and what your company wants to do. Some places fix the data at the source, for example.

You should have done enough analysis of your data that you have handled the vast majority of patterns, say 95% as a rough starting point.