DTrace is able to report on vfs information in FreeBSD (as well as a raft of other probes). DTrace is enabled by default in the 10 kernel so all you need to do is load the module then run the dtrace script.
Load the DTrace module
kldload dtraceall
Get the vfssnoop.d script from the FreeBSD forums. The whole thread is a treasure trove for disk monitoring.
Run it:
./vfssnoop.d
Watch the output for what is accessed:
# ./vfssnoop.d
cc1: warning: is shorter than expected
TIMESTAMP UID PID PROCESS CALL SIZE PATH/FILE
1555479476691083 0 1225 nfsd vop_getattr - /share/netboot
1555479478601010 0 1225 nfsd vop_inactive - /share/netboot
1555479482457241 0 1225 nfsd vop_getattr - /share/wpad.dat
1555480557262388 0 1432 cron vop_getattr - /var/cron/tabs
1555480557302178 0 1432 cron vop_inactive - /var/cron/tabs
1555480557336414 0 1432 cron vop_inactive - /etc
1555480557346224 0 1432 cron vop_getattr - /etc/crontab
top
to look for the proc that is eating your disk i/o? :top -m io -o total
: freebsd.org/cgi/…