Github user brandtg commented on the pull request:
https://github.com/apache/helix/pull/30#issuecomment-124923785
What's the specific use case for this feature? I'm imagining something
like, say, re-indexing multiple partitions of data. You want to just give it a
shot and let the chips fall as they may, then re-run the task on problem
partitions.
Maybe I'm not understanding it correctly, and there might be some gotcha
for this suggestion... But if you have a task that can be partitioned in to N
sequences of steps and require no synchronization between steps, why don't you
just model it as N tasks all with one partition (i.e. each has its own target
resource)?
Obviously this is more load on ZK, but I'm wondering if it'd be okay for
your use case. You could also leverage batch messaging to mitigate this I think.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
