Rob wrote:
David Lord <[email protected]> wrote:
David Woolley wrote:
David Lord wrote:
Systems seem heavily overloaded.
My 3G mobile broadband has seconds of latency whilst initiating
connections, that is how the system works, maybe 10 seconds, but
That's likely to be result of the power management design, not of overload.
No, connection is already established, PPP is already up and
routing all setup. Possibly ssh connected but no traffic.
He is correct, it is a power- and traffic management issue, not something
unique to the UK.
He's not correct
I ordered mobile broadband and order was rejected because
the network in this area is too congested and degrading
performance so I'll have to wait until next upgrade.
There are some polling rates in the 3G protocol that adapt to the actual
traffic. When a connection has been idle for long, it can take some time
to come back when you suddenly use it. When there is occasional traffic,
it will take around 400ms. Once there is lots of traffic (more than
a few packets per second), the roundtriptime will drop to 80-100ms and
stay there until there is less traffic.
E.g. pinging it at 1 second intervals will not keep it at the fast rate.
Pinging 5 times per second (ping -i 0.2) will.
I was seeing what I wrote and can't care less
about contra examples.
From today:
rtt min/avg/max/mdev(ms)
ping -c 10 -i 5 168/384/626/128
ping -c 10 -i 1 150/303/1044/265
ping -c 10 -i 0.2 382/484/758/102
ping -c 10 -i 0.1 368/629/1222/297
Proves your point?
David
_______________________________________________
questions mailing list
[email protected]
http://lists.ntp.org/listinfo/questions
