[Server-devel] XS 0.5 Second Attempt at eth1

[Martin Langhoff]

On Fri, Dec 5, 2008 at 12:06 AM, Jerry Vonau <jvonau at shaw.ca> wrote:
> Anna wrote:
>> So, what are the repercussions of this?
> Not sure... Martin?

Good sleuthing! Bonding-related errors was the last thing I'd
imagined, and with the e1000 driver too -- it's widely used and
generally well maintained.

Anna, what's your NIC? Ideally, we want to know the marketing
make/name/model and what lspci reports for it. Perhaps I can get my
hands on the same hw you have.

Googling a bit for e1000 and bonding to see if there are any known
issues I see quite a bit of traffic back in 2006/2007 with e1000
devices coming up / powering up late and confusing the bond drivers
and sometimes bad performance. But nothing current, it looks like it
got sorted out.

What's your other NIC? If you run xs-swapnics (and swap the cables
too), do things work better with the bonding configuration we're
shipping?

Also - I'm trying to think this through - I am fairly certain that one
of the test machines in Wellington has an e1000 and it worked
correctly for me (with a crossover cable to the AP).

cheers,



m
-- 
 martin.langhoff at gmail.com
 martin at laptop.org -- School Server Architect
 - ask interesting questions
 - don't get distracted with shiny stuff  - working code first
 - http://wiki.laptop.org/go/User:Martinlanghoff