abnormally high CPU load after zfs destroy

[Miroslav Lachman]

I am using ZFS pool (4x 3TB) as small backup storage. Backups are made 
by rsync and there are few snapshots. When I use "zfs destroy -r", there 
are high disk activity (it seems normal) but also high CPU load - 80+

The system did nothing at this time, just deleting old ZFS snapshot, so 
why is the load so high?

last pid: 90302;  load averages: 81.63, 43.60, 19.28                up 
43+03:33:04  19:56:16
36 processes:  1 running, 34 sleeping, 1 zombie
CPU:  0.0% user,  0.0% nice, 96.0% system,  0.0% interrupt,  4.0% idle
Mem: 5836K Active, 20M Inact, 4046M Wired, 755M Free
ARC: 1572M Total, 82M MFU, 1018M MRU, 24M Anon, 26M Header, 422M Other
Swap: 5120M Total, 17M Used, 5103M Free

   PID USERNAME     THR PRI NICE   SIZE    RES STATE   C   TIME    WCPU 
COMMAND
   592 root           1  20    0 26160K 18080K select  1   3:24   0.00% 
/usr/sbin/ntpd -g -c /etc/ntp.conf -p /var/run/ntpd.pid -f 
/var/db/ntpd.drift
   624 root           1  20    0 61224K  4204K select  1   2:20   0.00% 
/usr/sbin/sshd
   672 root           1  20    0 24136K  4592K select  0   1:10   0.00% 
sendmail: rejecting connections on daemon Daemon0: load average: 77 
(sendmail)
   655 root           1  20    0 25124K  4148K select  0   1:01   0.00% 
/usr/sbin/bsnmpd -p /var/run/snmpd.pid
   443 root           1  20    0 14512K  1760K select  1   0:43   0.00% 
/usr/sbin/syslogd -ss
   679 root           1  22    0 16612K   672K nanslp  1   0:32   0.00% 
/usr/sbin/cron -s
95875 xyz            1  20    0 65892K  4708K select  1   0:05   0.00% 
sshd: xyz @pts/0 (sshd)
   649 root           1  20    0 30704K  1432K nanslp  1   0:04   0.00% 
/usr/local/sbin/smartd -c /usr/local/etc/smartd.conf -p /var/run/smartd.pid
   352 root           1  20    0 13624K  1204K select  0   0:02   0.00% 
/sbin/devd
95873 root           1  20    0 65892K  4580K select  1   0:02   0.00% 
sshd: xyz  [priv] (sshd)
95912 root           1  20    0 25772K   652K pause   1   0:02   0.00% 
screen
   675 smmsp          1  20    0 24136K  1152K pause   0   0:01   0.00% 
sendmail: Queue runner at 00:30:00 for /var/spool/clientmqueue (sendmail)
89875 root           1  20    0 21940K  3188K CPU1    1   0:00   0.00% top
95914 root           1  20    0 23592K  3152K pause   0   0:00   0.00% 
-/bin/tcsh
95913 root           1  20    0 25772K  3192K select  1   0:00   0.00% 
screen
95895 root           1  52    0 23592K     0K pause   0   0:00   0.00% 
-su (<tcsh>)
95876 xyz            1  52    0 23592K     0K pause   1   0:00   0.00% 
-tcsh (<tcsh>)
95894 xyz            1  23    0 47740K     0K wait    0   0:00   0.00% 
/usr/bin/su - root (<su>)
90271 mrtg           1  52    0 17088K  2508K wait    1   0:00   0.00% 
/bin/sh ./local_iostat_disk.sh
89976 root           1  21    0 16612K  1704K piperd  1   0:00   0.00% 
cron: running job (cron)
90270 mrtg           1  52    0 17088K  2504K piperd  1   0:00   0.00% 
/bin/sh ./local_iostat_cpu.sh
   726 root           1  52    0 14508K  1700K ttyin   0   0:00   0.00% 
/usr/libexec/getty Pc ttyv0
   733 root           1  52    0 14508K  1700K ttyin   1   0:00   0.00% 
/usr/libexec/getty Pc ttyv7
   728 root           1  52    0 14508K  1700K ttyin   1   0:00   0.00% 
/usr/libexec/getty Pc ttyv2
   727 root           1  52    0 14508K  1700K ttyin   1   0:00   0.00% 
/usr/libexec/getty Pc ttyv1
   729 root           1  52    0 14508K  1700K ttyin   0   0:00   0.00% 
/usr/libexec/getty Pc ttyv3
   731 root           1  52    0 14508K  1700K ttyin   0   0:00   0.00% 
/usr/libexec/getty Pc ttyv5
   730 root           1  52    0 14508K  1700K ttyin   1   0:00   0.00% 
/usr/libexec/getty Pc ttyv4
   732 root           1  52    0 14508K  1700K ttyin   0   0:00   0.00% 
/usr/libexec/getty Pc ttyv6
90286 mrtg           1  52    0 18740K  2236K nanslp  1   0:00   0.00% 
iostat -w 250 -c 2 -x ada0 ada1 ada2 ada3
90283 mrtg           1  52    0 18740K  2228K nanslp  1   0:00   0.00% 
iostat -d -C -n 0 -w 240 -c 2
90287 mrtg           1  52    0 12356K  1952K piperd  1   0:00   0.00% 
tail -n 4
   137 root           1  52    0 12352K     0K pause   1   0:00   0.00% 
adjkerntz -i (<adjkerntz>)
90288 mrtg           1  52    0 17088K  2508K piperd  0   0:00   0.00% 
/bin/sh ./local_iostat_disk.sh
  3313 root           1  52    0 16728K  1884K select  1   0:00   0.00% 
/usr/sbin/moused -p /dev/ums0 -t auto -I /var/run/moused.ums0.pid



# uname -srmi
FreeBSD 10.2-RELEASE-p10 amd64 GENERIC

# grep CPU: /var/run/dmesg.boot
CPU: Intel(R) Pentium(R) Dual  CPU  E2160  @ 1.80GHz (1795.53-MHz 
K8-class CPU)


Miroslav Lachman