misordering packets [was [e2e] Link Aggregation]

[Reiner Ludwig]

At 16:34 13.05.2003, you wrote:
>[...] because most routers try to avoid misordering packets of a single 
>microflow [...]

That's interesting. Do you have a reference for that?

For example, "Packet reordering inside the Sprint backbone" by G. 
Iannaccone, S. Jaiswal, and C. Diot confirm what you say: "In contrast with 
what has been reported in previous studies, our analysis shows that 
reordering occurs in the backbone with a relatively low frequency."
(measurements from Aug. 2000)

An earlier study by J. C. R. Bennett, C. Partridge, and N. Shectman 
("Packet Reordering is Not Pathological Network Behavior") showed the 
opposite (at one particular point in the Internet, the MAE-East peering 
point): "We have found, however, that parallelism in Internet components 
and links is causing packet reordering under normal operation and that the 
incidence of packet reordering appears to be substantially higher than 
previousl reported."
(measurements from Dec. 1997 & Jan. 1998)

So, can we assume that packet reordering has become pathological network 
behavior in recent years? Does anybody know of more recent studies?

///Reiner