[
https://issues.apache.org/jira/browse/FLINK-7456?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16268209#comment-16268209
]
ASF GitHub Bot commented on FLINK-7456:
---------------------------------------
Github user zhijiangW commented on the issue:
https://github.com/apache/flink/pull/4552
@NicoK , thanks for focusing on the last PR.
I am supposed to divide this into two separate ones as you said. But I am
afraid some current tests may fail if I only modify and enable the credit-based
process on sender side, otherwise I may need to create a temporary handler in
parallel with `PartitionRequestQueue`.
As you said, the latter mainly replaces the `PartitionRequestClientHandler`
with the contents of `CreditBasedClientHandler`, also it replaces the
`CreditBasedClientHandler` class name in previous added tests and remove
previous temporary codes in `ResultPartitionType` to make fixed buffer pool
work.
If it still brings difficulty for your review, then I will try to divide it
then. : )
> Implement Netty sender incoming pipeline for credit-based
> ---------------------------------------------------------
>
> Key: FLINK-7456
> URL: https://issues.apache.org/jira/browse/FLINK-7456
> Project: Flink
> Issue Type: Sub-task
> Components: Network
> Reporter: zhijiang
> Assignee: zhijiang
> Fix For: 1.5.0
>
>
> This is a part of work for credit-based network flow control.
> On sender side, each subpartition view maintains an atomic integer
> {{currentCredit}} from receiver. Once receiving the messages of
> {{PartitionRequest}} and {{AddCredit}}, the {{currentCredit}} is added by
> deltas.
> Each view also maintains an atomic boolean field to mark it as registered
> available for transfer to make sure it is enqueued in handler only once. If
> the {{currentCredit}} increases from zero and there are available buffers in
> the subpartition, the corresponding view will be enqueued for transferring
> data.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
