I'm wondering when the GatheringByteChannel's write methods (taking in an
array of ByteBuffers) have advantages over the "regular" WritableByteChannel write methods.
I tried a test where I could use the regular vs. the gathering write method on a FileChannel, with approx 400KB/sec total in ByteBuffers of between 23-27 bytes in length in both cases. Gathering writes used an
array of 64. The regular method used up approx 12% of my CPU, and the gathering method used up approx 16% of my CPU (worse than the regular method!)
This tells me it's NOT useful to use gathering writes on a FileChannel around this range of operating parameters. Why would this be the case, and when would you ever use GatheringByteChannel? (on network I/O?)
Relevant differences here:
public void log(Queue<Packet> packets) throws IOException
{
if (this.gather)
{
int Nbuf = 64;
ByteBuffer[] bbufs = new ByteBuffer[Nbuf];
int i = 0;
Packet p;
while ((p = packets.poll()) != null)
{
bbufs[i++] = p.getBuffer();
if (i == Nbuf)
{
this.fc.write(bbufs);
i = 0;
}
}
if (i > 0)
{
this.fc.write(bbufs, 0, i);
}
}
else
{
Packet p;
while ((p = packets.poll()) != null)
{
this.fc.write(p.getBuffer());
}
}
}
