Hi!

On Thu, 26 Sep 2002, Jarkko Teppo wrote:

> bash-2.05# dmesg |grep "ex0: uplistptr was 0" |wc -l
>       61

> This machine has a huge load but it doesn't hang (at all), P3-400.

We've seen those "uplistptr" messages earlier, too. A quick count
shows six separate messages for just *one* of the ex0/ex1 between
2001/11/01 and 2002/08/26. But after 9/23 the machine has lost network
connectivity three times (twice yesterday, once today), each time
functionality was restored by logging in from the console, and checking
whether things work. Each time there has been the same "uplistptr" message
for *both* interfaces.

So, maybe the network hang and uplistptr messages are connected, maybe
not. But the network definately became unstable *after* upgrading (from
1.6E ca. August 8th) to 1.6H (ca. September 22/23).

I'd call this annoying behaviour for a machine that is supposed to forward
packets, but that's what you get with -current sometimes. :-/
(An optimist would say that the performance finally got so good, that
it starts to bring up errors on hardware. ;-)

I'm still interested in hearing any possible fixes for this. :)


Artsi
#######======------  http://www.selonen.org/arto/  --------========########
Everstinkuja 5 B 35                               Don't mind doing it.
FIN-02600 Espoo        arto@selonen.org         Don't mind not doing it.
Finland              tel +358 50 560 4826     Don't know anything about it.