Re: Frustrating network latency issue on NetBSD 6.1

["David Lord" <netbsd%lordynet.org@localhost>]

On 20 Jun 2013 at 14:45, Mayuresh wrote:

> On Wed, Jun 19, 2013 at 11:53:42AM -0400, Greg Troxel wrote:
> > Use ping, not ssh, to be able to have a more isolated test case and get
> > quantitative data.  In particular, separate loss from delay; both
> > present as delay over TCP.
> 64 bytes from 192.168.169.1: icmp_seq=0 ttl=64 time=6.781 ms
> 64 bytes from 192.168.169.1: icmp_seq=1 ttl=64 time=2.283 ms
.....
> 64 bytes from 192.168.169.1: icmp_seq=24 ttl=64 time=1.965 ms
> ^C
> ----router PING Statistics----
> 25 packets transmitted, 16 packets received, 36.0% packet loss
> round-trip min/avg/max/stddev = 1.937/2.639/6.781/1.253 ms
> 
> I think 36% is a large number for packet loss?

It is unless the link is very busy when icmp might be 
dropped.

What was your "ping" commandline?

Here with 100Mbit ethernet I get:
$ ping -c 100 me6000g
.....
100 packets transmitted, 100 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 0.327/0.470/0.527/0.035 ms

I could probably get some packet loss if I was to flood ping
and the link was not otherwise idle.

Traffic between pcs on my lan is mostly ipv6:
$ ping6 -c 100 me6000g
100 packets transmitted, 100 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 0.246/0.393/0.461/0.036 ms


David


> 
> How can this be checked further?
> 
> Actually my main router is not DD-WRT. It runs proprietary firmware (Lava
> W150). I was using WRT160NL DD-WRT more as a range extender. But to reduce
> the variables, currently I am directly connecting to the main router.
> where I don't have a way to examine the packets.
> 
> Mayuresh