[
https://issues.apache.org/jira/browse/FLINK-7406?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16186041#comment-16186041
]
ASF GitHub Bot commented on FLINK-7406:
---------------------------------------
Github user NicoK commented on a diff in the pull request:
https://github.com/apache/flink/pull/4509#discussion_r141902956
--- Diff:
flink-runtime/src/main/java/org/apache/flink/runtime/io/network/partition/consumer/RemoteInputChannel.java
---
@@ -70,6 +79,21 @@
*/
private int expectedSequenceNumber = 0;
+ /** The initial number of exclusive buffers assigned to this channel. */
+ private int initialCredit;
+
+ /** The current available buffers including both exclusive buffers and
requested floating buffers. */
+ private final ArrayDeque<Buffer> availableBuffers = new ArrayDeque<>();
+
+ /** The number of available buffers that have not been announced to the
producer yet. */
+ private final AtomicInteger unannouncedCredit = new AtomicInteger(0);
+
+ /** The number of unsent buffers in the producer's sub partition. */
+ private final AtomicInteger senderBacklog = new AtomicInteger(0);
+
+ /** The tag indicates whether this channel is waiting for additional
floating buffers from the buffer pool. */
+ private final AtomicBoolean isWaitingForFloatingBuffers = new
AtomicBoolean(false);
--- End diff --
Now seeing this in action: do we really need a `AtomicBoolean`? Or is a
`volatile boolean` enough? All uses except for `notifyBufferDestroyed()` (where
only a safety-check uses the value) are actually under a `synchronized
(availableBuffers)` block...in this case, you may also annotate the variable as
`@GuardedBy("availableBuffers")` for documentation.
> Implement Netty receiver incoming pipeline for credit-based
> -----------------------------------------------------------
>
> Key: FLINK-7406
> URL: https://issues.apache.org/jira/browse/FLINK-7406
> 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.
> Currently {{PartitionRequestClientHandler}} receives and reads
> {{BufferResponse}} from producer. It will request buffer from {{BufferPool}}
> for holding the message. If not got, the message is staged temporarily and
> {{autoread}} for channel is set false.
> For credit-based mode, {{PartitionRequestClientHandler}} can always get
> buffer from {{RemoteInputChannel}} for reading messages from producer.
> The related works are:
> * Add the backlog of producer in {{BufferResponse}} message structure
> * {{PartitionRequestClientHandler}} requests buffer from
> {{RemoteInputChannel}} directly
> * {{PartitionRequestClientHandler}} updates backlog for
> {{RemoteInputChannel}}, and it may trigger requests of floating buffers from
> {{BufferPool}}
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
