[c-nsp] Products for dealing with packet re-ordering?

[Saku Ytti]

On 11 January 2018 at 01:02, heasley <heas at shrubbery.net> wrote:

> is that entirely true? misorder is not (should not be) immediately
> interpretted as loss.  it does imply loss and might encite an
> implementation to ACK or duplicate ACK a preceding segment, but both
> sides should wait a reasonable time based on observed RTO for
> retransmission.
>
> so, if the implementation doesn't stink and the jitter isnt too bad ...

Implementation specific I'm sure. But at least the implementations
I've used while testing reordering, namely Linux, are entirely
intolerant to reorder, performance is immediately destroyed. I think
Window is same too. As long as you measure congestion on packet loss,
instead of say RTT time, the faster you react, the better performance
you have on lossy link (while entirely destroying ability to cope on
reordering links). But networks as a rule do not reorder, so most
people won't notice/complain when they lose reorder tolerance.

-- 
  ++ytti