[
https://issues.apache.org/jira/browse/FLINK-7416?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16201368#comment-16201368
]
ASF GitHub Bot commented on FLINK-7416:
---------------------------------------
Github user zhijiangW commented on the issue:
https://github.com/apache/flink/pull/4533
`notifyCreditAvailable` would be called by three logics in
`RemoteInputChannel`. They are `RemoteInputChannel#recycle`,
`RemoteInputChannel#notifyBufferAvailable` and
`RemoteInputChannel#onSenderBacklog` which are covered in previous PRs.
All the previous PRs are already merged into master except
[FLINK-7406](https://github.com/apache/flink/pull/4509).
In details, when the channel's credit is increased from zero, it will try
to notify the producer of it. For example:
1. Recycle the exclusive buffers to increase credit after record processed.
2. The buffer pool notifies the channel of available floating buffers to
increase credit
3. When receiving the backlog from producer, it may trigger to request
floating buffers from buffer pool. Credit is increased when returned floating
buffers from buffer pool.
> Implement Netty receiver outgoing pipeline for credit-based
> -----------------------------------------------------------
>
> Key: FLINK-7416
> URL: https://issues.apache.org/jira/browse/FLINK-7416
> Project: Flink
> Issue Type: Sub-task
> Components: Network
> Reporter: zhijiang
> Assignee: zhijiang
> Fix For: 1.4.0
>
>
> This is a part of work for credit-based network flow control.
> The related works are :
> * We define a new message called {{AddCredit}} to notify the incremental
> credit during data shuffle.
> * Whenever an {{InputChannel}}’s unannounced credit goes up from zero, the
> channel is enqueued in the pipeline.
> * Whenever the channel becomes writable, it takes the next {{InputChannel}}
> and sends its unannounced credit. The credit is reset to zero after each sent.
> * That way, messages are sent as often as the network has capacity and
> contain as much credit as available for the channel at that point in time.
> Otherwise, it would only add latency to the announcements and not increase
> throughput.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
