Github user koeninger commented on the pull request:
https://github.com/apache/spark/pull/3798#issuecomment-68804547
I'm hopeful that SPARK-4014 will be finalized soon, waiting on that before
doing the refactor for preferred locations. That will involve changing the
partition fields to add leader host and port, and not looking up leaders on
the executors until after a failure.
After that, I can take your suggestion and emulate the approach taken by
FileInputDStream to restore generatedRdds from a map of Time ->
Array[KafkaRddPartition]
On Mon, Jan 5, 2015 at 5:43 PM, Tathagata Das <[email protected]>
wrote:
> Any thoughts on my comments on driver fault-tolerance?
>
> â
> Reply to this email directly or view it on GitHub
> <https://github.com/apache/spark/pull/3798#issuecomment-68802901>.
>
---
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.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
