[lug] Temporary system hanging

[Daniel Webb]

On Mon, Apr 03, 2006 at 06:51:35PM -0600, Nate Duehr wrote:

> Daniel Webb wrote:
> >I have a strange situation where my system (I'll elaborate below) hangs for
> >2-10 seconds, then keeps going.  This can happen up to 5 times a minute, so
> >it's really getting on my nerves.
> 
> Anything in the system logs look new or interesting around the time of 
> the hangs?
> 
> (Sorry have to ask the obvious question, since you didn't say...)

Sigh.

Well, I could have sworn I checked, but sure enough (on the VNC server):

Apr  4 02:15:03 nerd kernel: NETDEV WATCHDOG: eth0: transmit timed out
Apr  4 02:15:03 nerd kernel: eth0: Transmit timed out, status 0000, PHY status 786d, resetting...
Apr  4 02:15:03 nerd kernel: eth0: link up, 100Mbps, full-duplex, lpa 0x45E1

There are tons of these in the logs, um, about as often as I've noticed the
hangs.  

I have recently (since these problem started) added a new machine to my
router, although the VNC system had no hardware changes and the router
(a Linksys WRT-54G) had no changes.  

The new machine is a laptop with a 10/100 PCMCIA network card.  Are there any
settings I can change to fix this?  For a different reason, I have actually
had two different PCMCIA network cards in while this problem was going on, so
I guess both caused the problem.  I was previously using a 10Mbit card for
several months without problems.

Looking online, I see people with similar problems being solved by using the
"noapic" kernel boot parameter.  Here goes nothing...