[K12OSN] LTSP Lab Slow to Boot

[Todd O'Bryan]

I was wrong. They're nVidia onboard NICs.

Here are the specs:

Tyan S2927 Thunder n3600B board
"Two nForce Pro 3600 integrated MAC with two Marvell 88E1116 single
port Gigabit Ethernet PHY"

The error we were getting was:

[38050.090000] eth0: too many iterations (21) in nv_nic_irq.
[38193.060000] eth0: too many iterations (21) in nv_nic_irq.

Again, this only happened with the 32-bit server kernel. 32-bit
desktop and 64-bit seemed fine.

Todd

On Sun, Sep 14, 2008 at 9:04 AM, Todd O'Bryan <toddobryan at gmail.com> wrote:
> On Sun, Sep 14, 2008 at 2:16 AM, Terrell Prude' Jr.
> <microman at cmosnetworks.com> wrote:
>> So, it sounds like you have a good setup here.
>>
>> --TP
>
> Thanks!
>
>> Interesting...what kind of onboard NICs are these?  And what kind of server
>> motherboard nowadays includes non-Linux-friendly NIC's?  I'd like to know,
>> so I can avoid them like the plague.
>
> They're actually Intel NICs and they seem to work fine with the
> regular 32-bit Linux kernel and the 64-bit kernel, but you get errors
> about too many connections with the 32-bit server kernel. I can't
> remember exactly what the problem was...one of my students who headed
> off to college was trying to track it down and said that the same
> error messages had been present back to 2006 without ever getting
> fixed. People on the mailing lists didn't seem to think it was
> something to worry about, but we were noticing hiccups on the LTSP
> network, probably because so many clients were trying to connect.
>
> I just emailed my student to see if he remembers the details and will
> send them along when he responds.
>
> Todd
>