Andrew Gillham wrote:
> This makes a good "denial of service" attack IMHO.  Your server happily
> runs along with 'NMBCLUSTERS=1024' for months, and suddenly you lose
access
> to it, have have to rebuild a kernel with 2048.  Oops, that isn't enough,
> lost access again, try 4096, oops, 8192, oops, 16384.  Whew!  Damn thing
> is "stable" again.

Is there any likelyhood that this can be addressed in some future version of
NetBSD?  I actually checked, and another NetBSD/Alpha box here (which does
DNS and sendmail) is running with 32768 -- before that it wasn't stable.

Is it not feasible for the kernel to recover from this, and get the network
running again?  Is this something inherent in the [Net]BSD kernel
architecture that cannot be cured?

Tom Haapanen
tomh@motorsport.com