Github user zhijiangW commented on a diff in the pull request:
https://github.com/apache/flink/pull/4499#discussion_r139596400
--- Diff:
flink-runtime/src/main/java/org/apache/flink/runtime/io/network/partition/consumer/RemoteInputChannel.java
---
@@ -183,18 +214,40 @@ void notifySubpartitionConsumed() {
}
/**
- * Releases all received buffers and closes the partition request
client.
+ * Releases all received and available buffers, closes the partition
request client.
*/
@Override
void releaseAllResources() throws IOException {
if (isReleased.compareAndSet(false, true)) {
+
+ final List<MemorySegment> recyclingSegments = new
ArrayList<>();
+
synchronized (receivedBuffers) {
Buffer buffer;
while ((buffer = receivedBuffers.poll()) !=
null) {
- buffer.recycle();
+ if (buffer.getRecycler() == this) {
+
recyclingSegments.add(buffer.getMemorySegment());
--- End diff --
I also considered this issue when implementation. The current way seems
more verbose than directly calling `recycle()`, but there are two advantages:
1. Exclusive buffers are recycled in batch which may be more performant as
you mentioned above.
2. If calling `recycle()` directly, it should add extra check `isReleased`
process outside the synchronized in `RemoteInputChannel#recycle` method. It
seems not elegant to do so. What do you think of this issue?
---
