[resolved] Resume on error

One Star

[resolved] Resume on error

Hi,
What is the best practice to create a resume-on-error process?
For example I have this:
tFileInputDelimited --> tMap --> tMySQLOutput.
And then the primary key field is a foreign and it's parent does not exist for one of the records in the inputfile.
Above is just an example, I am NOT looking for a specific solution on the tMysqlOutput component but a more generic solution which can also be used on other components.
Thanks

Accepted Solutions
Community Manager

Re: [resolved] Resume on error

Hi saukema
We usually finish the main processing in a child job, and use tRunJob to call the child job, uncheck the option 'die on error' on tRunJob. For example:
parent job:
tFileInputDelimited--main--tMap--main-->tFlowToIterate--iterate--tRunJobou
on tRunJob, uncheck the option 'die on error', so that the main job will continue to execute next record even though the child job has an error.
child job:
tFixedFlwInput--main--tMysqlOutput
----------------------------------------------------------
Talend | Data Agility for Modern Business

All Replies
Community Manager

Re: [resolved] Resume on error

Hi saukema
We usually finish the main processing in a child job, and use tRunJob to call the child job, uncheck the option 'die on error' on tRunJob. For example:
parent job:
tFileInputDelimited--main--tMap--main-->tFlowToIterate--iterate--tRunJobou
on tRunJob, uncheck the option 'die on error', so that the main job will continue to execute next record even though the child job has an error.
child job:
tFixedFlwInput--main--tMysqlOutput
----------------------------------------------------------
Talend | Data Agility for Modern Business