I recently upgraded our list server to 2.2.13pre3 (from 2.0.35) and have
started getting these kernel messages:
Sep 3 17:00:32 murphy kernel: eth0: Insufficient memory; nuking packet.
Sep 3 17:01:33 murphy last message repeated 1341 times
It seems from the source fro the de4x5 driver that it is trying to call
dev_alloc_skb which failes for some reason, so it drops the packet. This
is a very busy machine, it typically has several hundred active processes
(qmail) and handles about a half million outbound mails each day.
>From what little I know about the kernel I guess it probably is not
swapping agressively enough. The machine usualy has a free output of:
murphy{root}/home/jgg#free
total used free shared buffers cached
Mem: 63424 53724 9700 18060 9700 23548
-/+ buffers/cache: 20476 42948
Swap: 120924 2444 118480
But I think some of those procmail processes can really chew up ram fast.
So, is this a kernel flaw, or should I be looking at tuning one of the
parameters in /proc/ ?
Also, the 2.2 kernel has made a -huge- difference to the speed of the list
server, easialy twice as fast, possibly more!
Thanks,
Jason
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/