Subject: Re: -current (19991020 kernel, recent userland, i386) and sendto oddity
To: None <email@example.com>
From: Jarkko Teppo <firstname.lastname@example.org>
Date: 10/21/1999 11:04:03
From: Sean Doran <email@example.com>>
> traceroute: sendto: No buffer space available
AOL Me too!
Except this happens with 1.4.1 on pentium 200MHz with ~90MB
memory and two Cogent cards (de0, de1).
This is a bit problematic because I run gobs of mrtg processes and
NAT (cough!) and the mrtg-processes stay in the run que and in
the morning I might have something like >1000 processes.
Reboot helps but I'd say that's a suboptimal solution.
I've added MAXUSERS and pumped up NMBCLUSTERS which seems to have
> which persists until I do "ifconfig ep0 down; ifconfig ep0 up".
> any ideas what could be causing this, and how to fix it?