Re: netstat -m still reporting incorrect numbers

[Matthew Dillon <dillon@xxxxxxxxxxxxxxxxxxxx>]

:This is soon after the boot:
:
: > netstat -m
:257/1/7232 mbufs in use (current/peak/max):
:         257 mbufs allocated to data
:256/256/3616 mbuf clusters in use (current/peak/max)
:512 Kbytes allocated to network (5% of mb_map in use)
:0 requests for memory denied
:0 requests for memory delayed
:0 calls to protocol drain routines
: > Dec  4 15:35:53 dk su: tomaz to root on /dev/ttyp0
:
:While sending 5k HTTP requests per second...
: > netstat -m
:4294941443/4294963422/7232 mbufs in use (current/peak/max):
:         4294938235 mbufs allocated to data
:         3208 mbufs allocated to packet headers
:4462/4462/3616 mbuf clusters in use (current/peak/max)
:7955 Kbytes allocated to network (88% of mb_map in use)
:0 requests for memory denied
:0 requests for memory delayed
:0 calls to protocol drain routines
:
:a little later (still sending those 5k requests/s)
: > netstat -m
:4294903205/4294958335/7232 mbufs in use (current/peak/max):
:         4294895199 mbufs allocated to data
:         8006 mbufs allocated to packet headers
:9895/9895/3616 mbuf clusters in use (current/peak/max)
:17549 Kbytes allocated to network (194% of mb_map in use)
:0 requests for memory denied
:0 requests for memory delayed
:0 calls to protocol drain routines
:
:
:Using HEAD as of today....
:
:
:Toma¾

    Do you have any of the mpsafe sysctl's enabled?  If so turn them off
    and see if the problem still persists.  The mbuf statistics are not
    currently MP safe.  We have to make them per-cpu, but I'm not going to
    do it until afer the release.

					-Matt
					Matthew Dillon 
					<dillon@xxxxxxxxxxxxx>