On Wed, 5 Nov 2014, David H. Gutteridge wrote:
Hello,
I'm posting this to port-arm per msaitoh@'s request. I've noticed
when I use "netstat -m" on my Raspberry Pi, it doesn't seem to give
correct statistics. "netstat -inv" on the other hand, appears to be
correct. This isn't really an issue for me, but I don't know if it
indicates a larger concern with measuring certain statistics.
Anyway, on an amd64 machine that has almost no network activity, I
get output like this:
256 mbufs in use:
256 mbufs allocated to data
0 calls to protocol drain routines
Whereas, on my Raspberry Pi, with multiple ssh sessions open and
other network activity occurring, I get this:
1 mbufs in use:
1 mbufs allocated to packet headers
0 calls to protocol drain routines
Here's my rpi & imx23 olinuxino (7.0_BETA and -current). imx23 has longer
uptime here. rpi is fresh install from this morning (+pkgsrc fetch)
rpi$ uname -a && netstat -m
NetBSD rpi 7.0_BETA NetBSD 7.0_BETA (RPI) #1: Thu Nov 6 01:33:30 EET 2014 petri@babylon:/work/nb/obj/data/dev/code/NetBSD-7.0_BETA/src/sys/arch/evbarm/compile/RPI evbarm
1 mbufs in use:
1 mbufs allocated to packet headers
0 calls to protocol drain routines
oli$ uname -a && netstat -m
NetBSD oli 7.99.1 NetBSD 7.99.1 (IMX23_OLINUXINO) #0: Tue Oct 28 20:27:24 EET 2014 petri@babylon:/work/nb/obj/data/dev/code/NetBSD-current/src/sys/arch/evbarm/compile/IMX23_OLINUXINO evbarm
3 mbufs in use:
3 mbufs allocated to packet headers
48 calls to protocol drain routines