[
https://issues.apache.org/jira/browse/FLINK-7406?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16265239#comment-16265239
]
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_r152961281
--- Diff:
flink-runtime/src/main/java/org/apache/flink/runtime/io/network/partition/consumer/RemoteInputChannel.java
---
@@ -572,20 +560,53 @@ void addExclusiveBuffer(Buffer buffer) {
exclusiveBuffers.add(buffer);
}
- Buffer takeExclusiveBuffer() {
- return exclusiveBuffers.poll();
- }
-
void addFloatingBuffer(Buffer buffer) {
floatingBuffers.add(buffer);
}
- Buffer takeFloatingBuffer() {
- return floatingBuffers.poll();
+ /**
+ * Add the exclusive buffer into the queue, and recycle one
floating buffer if the
+ * number of available buffers in queue is more than required
amount.
+ *
+ * @param buffer The exclusive buffer of this channel.
+ * @return Whether to recycle one floating buffer.
+ */
+ boolean maintainTargetSize(Buffer buffer) {
--- End diff --
Sorry about the forth-and-back here, but thinking about the bug that you
fixed with the latest commit, it would be dangerous if we ever called
`maintainTargetSize()` without adding an exclusive buffer. What do you think
about my second suggestion instead, i.e. having a
```
/**
* Adds an exclusive buffer (back) into the queue and recycles
one floating buffer if the
* number of available buffers in queue is more than the
required amount.
*
* @param buffer
* the exclusive buffer to add
* @param numRequiredBuffers
* the number of required buffers
*
* @return how many buffers were added to the queue
*/
int addExclusiveBuffer(Buffer buffer, int numRequiredBuffers) {
exclusiveBuffers.add(buffer);
if (getAvailableBufferSize() > numRequiredBuffers) {
Buffer floatingBuffer = floatingBuffers.poll();
floatingBuffer.recycle();
return 0;
} else {
return 1;
}
}
```
(please note the changed return type which I think is more obvious than a
`boolean`)
> 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: 1.5.0
>
>
> 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)
