Re: odd ppp performance
On Sun, 23 Apr 2000 kmself@ix.netcom.com wrote:
> On Sun, Apr 23, 2000 at 11:51:35AM +0800, Lindsay Allen wrote:
> >
> > Greetings,
> >
> > I have been bothered for years with this problem and I still have no idea
> > on where to look for the cause.
> >
> > About once a fortnight I get very slow ppp performance and when I check
> > with ping I see the packets coming back in groups of ~8. Sometimes it
> > fixes itself but usually I have to hangup and redial. Here is an example.
> >
> > mentor:# ping t99
> > PING t99 t99.xxx.wa.edu.au (192.168.1.99): 56 data bytes
> > 64 bytes from 192.168.1.99: icmp_seq=0 ttl=254 time=7441.6 ms
> > 64 bytes from 192.168.1.99: icmp_seq=1 ttl=254 time=6459.8 ms
> > 64 bytes from 192.168.1.99: icmp_seq=2 ttl=254 time=5474.5 ms
> > 64 bytes from 192.168.1.99: icmp_seq=3 ttl=254 time=4517.5 ms
> > 64 bytes from 192.168.1.99: icmp_seq=4 ttl=254 time=3525.3 ms
> > 64 bytes from 192.168.1.99: icmp_seq=5 ttl=254 time=2532.6 ms
> > 64 bytes from 192.168.1.99: icmp_seq=6 ttl=254 time=1542.5 ms
> > 64 bytes from 192.168.1.99: icmp_seq=7 ttl=254 time=552.7 ms
> > 64 bytes from 192.168.1.99: icmp_seq=8 ttl=254 time=7996.1 ms
> > 64 bytes from 192.168.1.99: icmp_seq=9 ttl=254 time=7037.2 ms
> > 64 bytes from 192.168.1.99: icmp_seq=10 ttl=254 time=6061.9 ms
> > 64 bytes from 192.168.1.99: icmp_seq=11 ttl=254 time=5071.6 ms
> > 64 bytes from 192.168.1.99: icmp_seq=12 ttl=254 time=4077.2 ms
> > 64 bytes from 192.168.1.99: icmp_seq=13 ttl=254 time=3082.3 ms
> > 64 bytes from 192.168.1.99: icmp_seq=14 ttl=254 time=2108.0 ms
> > 64 bytes from 192.168.1.99: icmp_seq=15 ttl=254 time=1116.9 ms
> > 64 bytes from 192.168.1.99: icmp_seq=16 ttl=254 time=128.5 ms
> > 64 bytes from 192.168.1.99: icmp_seq=17 ttl=254 time=119.4 ms
> >
> > Can anyone throw any light on this weird behaviour, please?
>
> How long is your PPP connection up?
>
> I tend to leave my system online 24/7, and have noticed that after a
> time -- usually one the order of a week or two -- network performance
> degrades, though I hadn't thought to trace pings. What makes you say
> that they're coming in sets of 8?
>
> My solution is also to kill the connection and redial.
Yes, I do that too. But I want to know just what has gone wrong. I'm
like that. <g>
I leave my connections up 7/24 and it may perform just fine for maybe two
weeks. But then I get the above problem. (Local calls here are untimed.)
Why do I think the packets come in sets of 8? Take a look at the output.
The reply to ping 16 took 1 second longer than that for ping 17 and that
for packet 15 1 second longer again. Thus all 8 arrive as a burst. This
is confirmed by the modem lights.
This has occured with different Deb boxes and different ISPs. Seems to me
that it is either a ppp bug or a kernel bug but it has persisted over
different versions of both.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Lindsay Allen <allen@cleo.murdoch.edu.au> Perth, Western Australia
voice +61 8 9316 2486 32.0125S 115.8445E Debian Linux
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Reply to: