[
https://issues.jboss.org/browse/JGRP-1989?page=com.atlassian.jira.plugin....
]
Radim Vansa commented on JGRP-1989:
-----------------------------------
Now it really just passes the buffer into the queue, but in the past (looking into git
history before the NIO stuff) you've copied the buffer there:
{code:java}
public void send(byte[] data, int offset, int length) throws Exception {
if (sender != null) {
// we need to copy the byte[] buffer here because the original buffer might
get changed meanwhile
byte[] tmp = new byte[length];
System.arraycopy(data, offset, tmp, 0, length);
sender.addToQueue(tmp);
}
else
_send(data, offset, length, true, true);
}
{code}
So the current state when you allocate the buffer anew is not that worse after all
(besides allocating more than you need), because it just moved the invocation to bundler.
Bundlers: reuse send buffer when transport == sync.
---------------------------------------------------
Key: JGRP-1989
URL:
https://issues.jboss.org/browse/JGRP-1989
Project: JGroups
Issue Type: Enhancement
Reporter: Bela Ban
Assignee: Bela Ban
Fix For: 3.6.7
With the addition of {{TCP_NIO2}}, all bundlers now create new send buffers for every
message (or message list). This generates a lot of memory allocations, perhaps it is
better to revert this change for *synchronous transports* such as {{UDP}} and {{TCP}}, and
still create new buffers for *asynchronous transports* such as {{TCP_NIO2}}.
Synchronous transports guarantee a message has been put on the wire when {{TP.send()}}
returns, whereas asynchronous transports may only have completed a partial write (so we
cannot reuse the buffer).
The code in the bundler should check for this, and copy if async or not copy if sync.
Whether or not a transport is sync is determined by a new abstract method that needs to
be overridden by every transport.
--
This message was sent by Atlassian JIRA
(v6.4.11#64026)