[c-nsp] The myths of autonegotiate vs forced (was: full duplex mismatch speed - dynamips)

[Geoffrey Pendery]

Yes.
We recently rolled out new switches in a couple sites (4507-R with
Sup6E and 10/100 blades) and were checking to see if any clients came
up wrong.  We panicked when we saw tons of clients coming up 10/full,
10/half, and 100/half.  All over the map.

Turns out all the workstations were powered off during the install,
and it was just the WoL drivers that were wonky.
Once booted everything auto'ed up properly.

-Geoff


On Fri, Aug 20, 2010 at 11:28 AM, Jeff Kell <jeff-kell at utc.edu> wrote:
>  While we're on the autoneg subject...
>
> Has anyone else run into cases of newer computers that support a powersave and/or
> "wake-on-lan" state where in that instance the interface goes into a 10Mb/full connection?
>
> Drives us nuts with the growing "green" initiatives...
>
> Computers that do that go into some sort of zombie state that doesn't quite autoneg
> properly, our switches (Cisco and Procurve) go to 10/full, but the device doesn't seem
> to quite be full, and the ports will generate lots of up/downs and errors while in that
> state.
>
> If you're doing any sort of monitoring that involves trapping mac-address changes or
> linkup/linkdowns, they'll drive your monitors crazy too...
>
> When they're "working" they'll do gig/full, but when in this zombie state, they autoneg
> down to 10/full
>
> Jeff
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
>