-
Notifications
You must be signed in to change notification settings - Fork 109
Source, map, collection, sink for bounded data streams in Flink #661
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: main
Are you sure you want to change the base?
Conversation
|
yes, pls start in our dev mailing list. |
|
Please LMK if I need to deprecate any parts of the old DataSet API. also I'm unsure whether its a smart idea to implement |
|
@juripetersen @zkaoudi @novatechflow Review when you have time pls |
| private DataStream<?> dataStream; | ||
|
|
||
| // TODO: this.size is currently always 0 | ||
| private long size; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Does this have any effects?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Leftovers from the old DataSetChannel implementation the user could theoretically extend the DataStreamChannel and provide their own, but its such a niche situation I'm fine with removing it.
| @@ -56,6 +56,12 @@ public class FlinkExecutor extends PushExecutorTemplate { | |||
| */ | |||
| public ExecutionEnvironment fee; | |||
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think ideally we just remove the old Env and also DataSet with it, so that the end user of wayang doesn't even notice this change (other than a few operators missing maybe)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Per discussion in the dev-list I've delegated this to the configuration.
| import org.apache.wayang.flink.operators.FlinkBoundedTextFileSource; | ||
| import org.apache.wayang.flink.platform.FlinkPlatform; | ||
|
|
||
| public class BoundedTextFileSourceMapping implements Mapping { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
As everything is bounded as of now in Wayang, I would just call this TextFileSource and replace TextFileSource from Flink with it.
We can later add a StreamedTextFileSource.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
See comment above.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Nevertheless since our discussion on continuous streams hasn't concluded either I still propose we keep the bounded semantic as a way to leave the door open for continuous sources later.
| /** | ||
| * Mapping from {@link JoinOperator} to {@link FlinkDataStreamJoinOperator}. | ||
| */ | ||
| public class StreamedJoinMapping implements Mapping { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Same goes for this operator. I think we should just remove the DataSet ones and then replace them with the DataStream.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
See comment above.
juripetersen
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I left some comments, but we should just discuss if that's the way to go.
|
I guess we just need more activity in the discussion on the mailing list. |
lets have a discussion about how we are going to handle Continuous sources e.g. in flink. And how we are going to integrate this vs the old data set implementation.