[c-nsp] Cisco IOS ping utility reports lower RTT than possible

adamv0025 at netconsultings.com adamv0025 at netconsultings.com
Mon May 6 03:46:13 EDT 2019


> Martin T
> Sent: Friday, May 3, 2019 6:00 PM
> 
> Hi Adam,
> 
> > So how should I interpret the above tcpdump output please?
> Timestamp at the beginning of each line is simply a delta in micro-second
> resolution between current and previous line. So first tcpdump captured an
> ICMP "echo request" message, then 10201 µs later it captured ICMP "echo
> reply" message, then 431 µs later it captured next ICMP "echo request" and
> then 10165 µs later it captured reply for seq 1, etc.
> 
Hi Martyn,
Right so you now have 2 independent measurements all showing ~10+ ms latency (3 if you count juniper-mx). 
So where is the cisco getting the 8-12 spread is a mystery indeed. It's clearly a bug. Maybe try a different code on the cisco?

Or maybe the cisco box is deeper in gravitational well than the rest of the setup -resulting in its clocks to tick slower according to the general relativity :)

adam 



More information about the cisco-nsp mailing list