Subject: strange netstat -m output
To: None <netbsd-users@netbsd.org>
From: Emre Yildirim <emre.yildirim@us.army.mil>
List: netbsd-users
Date: 04/11/2001 11:17:46
Hi,

I have two NetBSD servers.  Server A runs NetBSD 1.5S, and server B 
runs NetBSD 1.5T.  Both have different kernels, same networking 
options except NMBCLUSTERS.

On server A, with options NMBCLUSTERS=8192 in the kernel I get:

# netstat -m
129 mbufs in use:
        128 mbufs allocated to data
        1 mbufs allocated to packet headers
128/768 mapped pages in use
29748 Kbytes allocated to network (0% in use)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1576 requests for memory denied
0 requests for memory delayed
1576 calls to protocol drain routines


On server B, with options NMBCLUSTERS=32768 in the kernel I get:

# netstat -m
130 mbufs in use:
        128 mbufs allocated to data
        2 mbufs allocated to packet headers
128/218 mapped pages in use
520 Kbytes allocated to network (55% in use)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
0 requests for memory denied
0 requests for memory delayed
0 calls to protocol drain routines


So why does server A have more memory allocated to the network, even 
though it has a lower NMBCLUSTERS value?  Does that have to do with 
anything?  Is there a way I can increase memory allocated to the 
network on server B without restarting (i.e. with sysctl maybe)?

Thanks for any pointers/help