[e2e] ResiliencyToward Packet Misordering

[Cottrell, Les]

We made a separate measure of re-ordering for 1 second separated pings in December 1998. There were about 220 paths in about 70 countries. Pretty much a subset of the for the measurement made in August 2000. We did not look at reorders for the 1 second separated pings for the August 2000 measurements.

Monitoring out-of-order packets for two weeks in December 1998, in an approximate total of 224,448 sets of pings (each set was 10 pings separated by 1 second) sent from SLAC in that time, 24 had out of order packets, roughly 1 in 10,000. Of the 24, 5 came in a short burst on December 16th between 04:00 and 10:00 (PST) all from sarka.fzu.cz. 4 more occurences to ftp.physics.carleton.ca happened on December 10,11,12 and 15. 3 Other sites (ntp.dnai.com, mnhepw.hep.umn.edu, uae6.ciemat.es) had 3 occurences each, there were two occurences to fnal.fnal.gov, and 4 other sites had one incident each. 

In all cases the out-of-order packet was due to an extraordinarily long response time. Some were long for all the pings, some had just the out-of-order ping taking up to 68 times as long to return as the minimum time for that sample. 


> -----Original Message-----
> From: Sharad Jaiswal [mailto:sharad at cs.umass.edu] 
> Sent: Thursday, August 01, 2002 7:42 AM
> To: Cottrell, Les
> Cc: end2end-interest at postel.org
> Subject: RE: [e2e] ResiliencyToward Packet Misordering
> 
> 
> Did you see any difference in the reordering numbers between 
> experiments done with back-back pings, and those done with a 
> separation of 1 second?
> 
> Sharad.
> 
> ____________________________________________________________________
> 
> sharad jaiswal                     http://www.cs.umass.edu/~sharad
> ____________________________________________________________________
> 
> On Wed, 31 Jul 2002, Cottrell, Les wrote:
> 
> > For each host in turn:
> > 1. We sent 5 56 bytes pings separated by 1 second to each 
> host. This 
> > was to prime caches etc. We looked to see whether any of 
> these pings 
> > were successful. If not we noted it and went on to the 
> following host 
> > (i.e. we skipped steps 2 and 3 below). 2. We then sent 50 
> back to back 
> > 56 byte pings using the preload option of the NIKHEF ping 
> and analyzed 
> > the response to extract and record the minimum/average/maximum 
> > response time, the loss, the inter packet separation, the number of 
> > packets out of order (a packet was considered out of order if its 
> > sequence number was less than the previous successfully received 
> > packet) and any pathologies such as duplicate packets. 3. 
> Step 2 was 
> > then repeated for 50 56 byte ping packets sent with a 
> separation of 1 
> > second and a timeout of 20 seconds. We repeated the 
> measurements with 
> > 1400 byte packets.
> >
> > -----Original Message-----
> > From: Vern Paxson [mailto:vern at icir.org]
> > Sent: Wednesday, July 31, 2002 6:55 PM
> > To: Cottrell, Les
> > Cc: 'Craig Partridge'; tvpoh at essex.ac.uk; 
> end2end-interest at postel.org
> > Subject: Re: [e2e] ResiliencyToward Packet Misordering
> >
> >
> > > Roughly 25% of the hosts monitored exhibit reordering ...
> >
> > What was the spacing between the packets you sent, and how many did 
> > you send?  A key point with reordering is that the rate is 
> a function 
> > of interpacket spacing.
> >
> > 		Vern
> >
> >
> 
>