Re: 1Gb/s network issues - was: Re: OT: GPS18x LVC failure

David J Taylor <david-taylor@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx> wrote:

"Unruh" <unruh-spam@xxxxxxxxxxxxxx> wrote in message
news:QsBAm.47742$Db2.29202@xxxxxxxxxxx
[]
Actually those Gb ethernet switches seems to have some really bad and
variable
latencies. My system, before they installed Gb switches, had a
consistant
roundtrip of 140 usec and gave better than 20usec jitter (with 100Mb
switches). Since those new switches
were installed things have gone to hell-- variable roundtrips ( 140usec,
300usec
and up to 1 sec roundtrip) and much much worse jitter. NOt sure where
the problem
is, but it sure is terrible.
Doing a plot of roundtrip vs offset it is definitely one way delay.

Most interesting, Bill. Mine is a relatively cheap and cheerful device
from Netgear (GS608):

http://www.netgear.co.uk/ethernet_network_switch_gs605.php

so whether being very simple and unmanaged is likely to mean that it's
performance is better or worse I leave to speculation or measurement!

Note that the fact that the jitter starts when the switch is replaced
does not mean that the problem must be in the switch.

It may well be that the interrupt handling on the network card is different
between 100 Mbps and 1 Gbps mode. It could be that interrupt coalescing
is enabled when the link is detected as 1 Gbps. So when you replace your
10/100 switch by a gigabit switch, you suddenly see a problem. But that
problem does not originate in the switch but in the network card/driver.

When a switch introduces a large roundtriptime or jitter it must mean
it has a lot of buffer capacity available. It is not so likely that a
manufacturer of a cheap device includes this fast buffer memory by
accident, I think.
.

Relevant Pages

  • 1Gb/s network issues - was: Re: OT: GPS18x LVC failure
    ... My system, before they installed Gb switches, had a consistant ... were installed things have gone to hell-- variable roundtrips (140usec, ... Doing a plot of roundtrip vs offset it is definitely one way delay. ... jitter) was using the network controllers built into the PCs. ...
    (comp.protocols.time.ntp)
  • Re: Bizzare behaviour from SG/UC3525
    ... >> jitter it's affected by my hand capacitance on the switch that I ... > Hand capacity effects around the Shutdown input? ... It seemed to stop the 'jitter' on the trailing high side edge but then I ... the output devices. ...
    (sci.electronics.design)