From: adams@achibm5.chemie.uni-karlsruhe.de (Klaus-Georg Adams) Subject: Re: net-2 eth0 receiver overrun (can't find thread) Date: 1 Aug 1993 17:17:25 GMT
In article <1993Jul28.172631.5757@ringer.cs.utsa.edu>, djimenez@ringer.cs.utsa.edu (Daniel Jimenez) writes:
|> Ok, I think it's safe to say that there is definitely a memory leak in
|> Linux 0.99pl11. Many people seem to be posting the same kinds of problems:
|> after many hours of uptime, the system grinds to a halt "out of memory."
|>
|> I tried an experiment on my system: I killed all the daemons except init,
|> then watched /proc/meminfo (using a sleep; cat shell script) for a while.
|> Every once in a while, the available memory would decrease, even though no
|> one else was logged on and no other processes were running (the `buffers'
|> portion of the information stayed the same).
|>
|> It seems everyone who has posted has the net-2 stuff compiled in (please
|> correct me if I'm wrong). What's going on?
|> --
|> Daniel Jimenez djimenez@ringer.cs.utsa.edu
|> "I've so much music in my head" -- Maurice Ravel, shortly before his death.
|> " " -- John Cage
I had the same problem (receiver-overrun etc...). It went away the moment, I changed my
netmask from 255.255.255.0 to 255.255.0.0 and my network from 129.13.108.0 to 129.13.0.0.
From that time on, it has never occured again.
Configuration: Linux 0.9.9pl11
net-2
SMC Elite 16 (WD8013-compatible)
I heavily use nfs and X-Windows over the network, and before the change, my machine would
just die after about 30 min of work.
BTW, my /proc/meminfo is stable as a rock, so I don't think, the problem is with the kernel.
Klaus-Georg Adams adams@achibm1.chemie.uni-karlsruhe.de