Re: Weird ping statistics...

[George Harvey <fr30%dial.pipex.com@localhost>]

On Sat, 15 Jan 2011 18:04:33 +0100
Johnny Billquist <bqt%softjar.se@localhost> wrote:

> I just repeated this test on a VAX-4000/90 and a MacBook PRO, and I can 
> observe the same thing, although with less frequency. But occasional 
> pings from the VAX to the Mac also get about 200 ms time, but not in the 
> other direction. (With some load on the VAX.)

I tried your ping test on a couple of my VAX boxes, both running 5.1
GENERIC, and see similar symptoms. The machines were a 3100/10e and a
3100/95, both running a C compile at the same time. The 95 showed
occasional jumps to 200ms but on the, much slower, 10e the jumps were
mostly to around 100-130ms with just a few in the 200-300ms range.

I also tried it on a Cobalt RaQ 2 running 5.0.2 and that produced a
199ms result as well so it's not just a VAX effect.

Short excerpts from the logs are included below.

George

== ping test on Microvax 3100/95 ==
64 bytes from 10.0.1.164: icmp_seq=591 ttl=255 time=1.680 ms
64 bytes from 10.0.1.164: icmp_seq=592 ttl=255 time=1.852 ms
64 bytes from 10.0.1.164: icmp_seq=593 ttl=255 time=1.808 ms
64 bytes from 10.0.1.164: icmp_seq=594 ttl=255 time=1.806 ms
64 bytes from 10.0.1.164: icmp_seq=595 ttl=255 time=200.686 ms
64 bytes from 10.0.1.164: icmp_seq=596 ttl=255 time=1.803 ms
64 bytes from 10.0.1.164: icmp_seq=597 ttl=255 time=1.831 ms
64 bytes from 10.0.1.164: icmp_seq=598 ttl=255 time=1.723 ms
64 bytes from 10.0.1.164: icmp_seq=599 ttl=255 time=1.848 ms

== ping test on Microvax 3100/10e ==
64 bytes from 10.0.1.164: icmp_seq=576 ttl=255 time=10.000 ms
64 bytes from 10.0.1.164: icmp_seq=577 ttl=255 time=10.000 ms
64 bytes from 10.0.1.164: icmp_seq=578 ttl=255 time=20.000 ms
64 bytes from 10.0.1.164: icmp_seq=579 ttl=255 time=20.000 ms
64 bytes from 10.0.1.164: icmp_seq=580 ttl=255 time=20.000 ms
64 bytes from 10.0.1.164: icmp_seq=581 ttl=255 time=120.000 ms
64 bytes from 10.0.1.164: icmp_seq=582 ttl=255 time=20.000 ms
64 bytes from 10.0.1.164: icmp_seq=583 ttl=255 time=10.000 ms
64 bytes from 10.0.1.164: icmp_seq=584 ttl=255 time=110.000 ms
64 bytes from 10.0.1.164: icmp_seq=585 ttl=255 time=10.000 ms
64 bytes from 10.0.1.164: icmp_seq=586 ttl=255 time=10.000 ms
... 
64 bytes from 10.0.1.164: icmp_seq=1396 ttl=255 time=10.000 ms
64 bytes from 10.0.1.164: icmp_seq=1397 ttl=255 time=10.000 ms
64 bytes from 10.0.1.164: icmp_seq=1398 ttl=255 time=10.000 ms
64 bytes from 10.0.1.164: icmp_seq=1399 ttl=255 time=150.000 ms
64 bytes from 10.0.1.164: icmp_seq=1400 ttl=255 time=20.000 ms
64 bytes from 10.0.1.164: icmp_seq=1401 ttl=255 time=10.000 ms
64 bytes from 10.0.1.164: icmp_seq=1402 ttl=255 time=10.000 ms
64 bytes from 10.0.1.164: icmp_seq=1403 ttl=255 time=50.000 ms
64 bytes from 10.0.1.164: icmp_seq=1404 ttl=255 time=260.000 ms
64 bytes from 10.0.1.164: icmp_seq=1405 ttl=255 time=10.000 ms
64 bytes from 10.0.1.164: icmp_seq=1406 ttl=255 time=10.000 ms
64 bytes from 10.0.1.164: icmp_seq=1407 ttl=255 time=10.000 ms

== ping text on Cobalt RaQ 2 ==
64 bytes from 10.0.1.164: icmp_seq=241 ttl=255 time=2.064 ms
64 bytes from 10.0.1.164: icmp_seq=242 ttl=255 time=2.015 ms
64 bytes from 10.0.1.164: icmp_seq=243 ttl=255 time=1.993 ms
64 bytes from 10.0.1.164: icmp_seq=244 ttl=255 time=2.033 ms
64 bytes from 10.0.1.164: icmp_seq=245 ttl=255 time=199.899 ms
64 bytes from 10.0.1.164: icmp_seq=246 ttl=255 time=2.004 ms
64 bytes from 10.0.1.164: icmp_seq=247 ttl=255 time=2.014 ms
64 bytes from 10.0.1.164: icmp_seq=248 ttl=255 time=2.021 ms
64 bytes from 10.0.1.164: icmp_seq=249 ttl=255 time=1.753 ms