A customer called in to complain that his cvc daemon was using 12.5% CPU time
on his 8-CPU machine, which amounts to a whole CPU. We compared this with our
lab machine and found that our cvc daemon was using 25% CPU time on our 4 CPU
machine.

We confirmed this as follows:

  # /usr/ucb/ps -auxww
  USER       PID %CPU %MEM   SZ  RSS TT       S    START  TIME COMMAND
  root         8 25.0  0.1  592  408 ?        O   Aug 20 39102:40 /sbin/cvcd
  ...
  # mpstat 5
  ...
  CPU minf mjf xcal  intr ithr  csw icsw migr smtx  srw syscl  usr sys  wt idl
    0    0   0    0   203    3   20    0    0    0    0     0    0   0   0 100
    1    0   0    0     2    0    2    2    0    0    0     0  100   0   0   0
    2    0   0    0     0    0   28    0    0    0    0     7    0   0   0 100
    3    1   0   51     0    0   28    0    0    0    0     2    0   0   0 100

After looking at bugID# 4171670, we decided to check /etc/nsswitch.conf and
/etc/hosts.

We fixed this by changing his /etc/nsswitch.conf file's "hosts" line to read:
  hosts:	files nisplus dns

and then stopping & restarting the cvc daemon:
  # /etc/init.d/cvc stop
  stopping Network Console
  # /etc/init.d/cvc start
  starting Network Console


The CPU usage of the cvc daemon is now restored to a more normal level:
  # /usr/ucb/ps -auxww
  USER       PID %CPU %MEM   SZ  RSS TT       S    START  TIME COMMAND
  root     20353  0.3  0.5 2136 2064 pts/0    O 17:00:01  0:00 /usr/ucb/ps -auxww
  root     20351  0.1  0.2  936  712 ?        S 16:59:54  0:00 /sbin/cvcd
  ...

I've seen the cahnges to /etc/nsswitch.conf mentioned in the "UE10000 System
Administration" (course# ES400) student guide, however this needs to be included
in the "Sun Enterprise 10000 Hardware Installation and De-Installation Guide"
(PN# 805-4651-11) section 6.3 at the bottom of page 6-14.