[gPXE] Strange ARP problem with gpxe 1.0.0+

[Ben Derstine]

I agree. I have it working with the sleep command for now and will switch over (pardon the pun) our 2960s as I can do this after hours.   For the moment the sleep command gives me a quick around until all our switches in all our locations can be reconfigured. 

Ben

-----Original Message-----
From: Stefan Hajnoczi [mailto:stefanha at gmail.com] 
Sent: Tuesday, May 18, 2010 2:50 PM
To: Ben Derstine
Cc: Clive Eisen; gpxe at etherboot.org
Subject: Re: [gPXE] Strange ARP problem with gpxe 1.0.0+

On Tue, May 18, 2010 at 5:16 PM, Ben Derstine <BLD17 at sa.psu.edu> wrote:
> I didn't but putting on seems to have solved problem- thanks.  I guess my
> gpxe script is trying to hit the network before the switch gets done with
> spanning-tree after the interface first comes up (it takes about 30 seconds
> on these switches.)  I'm going to see if I can put a timing delay in my gpxe
> script to allow for this.

Why not use portfast mode?  The purpose of portfast is to allow leaf
nodes to access the network quickly without waiting for STP.

Sleep will just add an artificial delay that slows down the boot.
There is a sleep command in gPXE but the solution is really to
configure the switches rather than adding delays in the boot process.

Stefan