[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: nmbclust and netstat -m



I can't speak for the original problem, but the discrepancy on what
should be done with NMBCLUSTERS is rather annoying.

I figured out why I was running out of NMBCLUSTERS, basically mail came
in faster than my server could deliver it and it bumped into the mclpool
max.

What exactly am I (or him) suppose to do other than raise NMBCLUSTERS?
The kernel message explicitly says this is what should be done.

Theo says someone should sit down and fix these things, but it seems
that the only solution for me (knowing jack about kernel programming)
would be to get -crazy- fast disks, or start splitting the load on a
server that, for the most part, is able to deal with its traffic load
quite handily.


I'm really not technical at all when it comes to kernel related things
like this, but I've been running fine with raised NMBCLUSTERS. Looking
at FreeBSD, it seems to have an insanely high number for nmbclust by
default; and of course it has no problems.




Chris


On Wed, 2004-04-28 at 08:47, Peter Hessler wrote:
> Don't do this.  You are only creating problems for yourself.  Read the
> archives.
> 
> On Wed, 28 Apr 2004 10:07:56 -0400
> "Todd Boyer" <tboyer_(_at_)_autumntech_(_dot_)_com> wrote:
> 
> :changed nmbclust from 2048 to 4096



Visit your host, monkey.org