On Fri, 12 Jul 2002, David Laight wrote:

> > There are two tests that are illustrative. In both, the tester sends a NOP
> > PDU (48 bytes) with 4k of attached data. In one case, the target only
> > sends a reply (48 bytes). In the other, it sends a reply and a copy of the
> > 4k of data. The first test flies along at line speed. The second one gets
> > hit by the .2 second delay.
>
> Stinks of a single bit heuristic being used to determine whether
> to wait for more data.  The short write gets sent immediately
> because the previous block was sent on timeout, and the long block
> gets sat on because it should (or could) have been usefully
> attached to the previous block.

I should probably mention I'm setting TCP_NODELAY to one on this, so I'd
expect any such heuristics to be deactivated. :-)

Also, setting 'sysctl -w net.inet.tcp.ack_on_push=1' makes the problem go
away, but at an increased packet flow.

Take care,

Bill