Github user uce closed the pull request at:
https://github.com/apache/flink/pull/471
---
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
Github user uce commented on the pull request:
https://github.com/apache/flink/pull/471#issuecomment-83049370
Thanks for the comments. I'll address the generateSequence comment and open
an issue for the optimizer deadlock detection and pipeline breaking stuff after
merging this.
---
Github user StephanEwen commented on the pull request:
https://github.com/apache/flink/pull/471#issuecomment-82555635
Once this is in, we can start removing the deadlock detection in the
optimizer and the pipeline breaking caches.
---
If your project is set up for it, you can reply t
Github user StephanEwen commented on the pull request:
https://github.com/apache/flink/pull/471#issuecomment-82554365
Look good, nice test coverage and fits very well with the recent execution
mode changes.
Only downside: This pull request does contains many cases where only
Github user uce commented on the pull request:
https://github.com/apache/flink/pull/471#issuecomment-82479502
I've rebased this on the current master including the `ExecutionMode`
changes introduced by Stephan.
This allows you to set the type of produced runtime result partiti
Github user uce commented on the pull request:
https://github.com/apache/flink/pull/471#issuecomment-79017193
I've rebased this on the latest master and set the default I/O mode to
synchronous, i.e. we currently use the simpler synchronous spilled subpartition
view when consuming inte
Github user uce commented on the pull request:
https://github.com/apache/flink/pull/471#issuecomment-78481915
PS: where we wouldn't need the asynchronous implementations are for local
reads. There it should be perfectly fine to just synchronously read the spilled
partitions.
---
If
Github user uce commented on the pull request:
https://github.com/apache/flink/pull/471#issuecomment-78479167
The root "cause" of all asynchronous operations is that TCP connections are
shared among multiple logical channels, which are handled by a fixed number of
network I/O threads.
Github user StephanEwen commented on the pull request:
https://github.com/apache/flink/pull/471#issuecomment-78474538
Concerning the questions:
1. I think deploying after all blocking producers are finished is what we
should go for as a start. It is also what people would exp
GitHub user uce opened a pull request:
https://github.com/apache/flink/pull/471
[FLINK-1350] [runtime] Add blocking result partition variant
- **Renames** runtime intermediate result classes (after an offline
discussion with @StephanEwen I realized that calling subpartitions *queue*
10 matches
Mail list logo