topic Multi-task Job Run starting point in Data Engineering

Multi-task Job Run starting point

Anonymous — Wed, 20 Oct 2021 15:19:15 GMT

Hi community!

I would like to know if it is possible to start a Multi-task Job Run from and specific task. The use case is as follows:

I have a 17 tasks Job
A task in the middle, let's say a task after 2 dependencies, fails
I found the error and now it is good to go
I want to restart the Job Run from the failed task and not all over again

Is it possible to accomplish the last bullet point or I'll have to run the whole job all over again?

Re: Multi-task Job Run starting point

Dan_Z — Wed, 20 Oct 2021 15:45:56 GMT

Hey @Esteban Sierra-Munera , this is actually an area of active development for the Jobs team at Databricks. We expect this functionality in Private Preview late this year or early next year. It's going to be really slick. In the mean time, what you can do is:

duplicate the job
delete the tasks that have succeeded in the original job
run that

Re: Multi-task Job Run starting point

BilalAslamDbrx — Fri, 29 Oct 2021 13:41:54 GMT

+1 to what @Dan Zafar said. We're working **** ** this. Looking forward to bring this to you in the near future.