[CentOS] Strange ntp behaviour?

Mon Nov 22 00:51:41 UTC 2010
Robert Moskowitz <rgm at htt-consult.com>

The following is from my logwatch email:

  --------------------- XNTPD Begin ------------------------


  XNTPD Killed: 7 Time(s)

  XNTPD Started: 8 Time(s)

  Total interfaces 8 (non-local: 4)

  Total synchronizations 6 (hosts: 1)

  Errors
    getaddrinfo: "0.centos.pool.ntp.org" invalid host address, ignored: 2 time(s)
    getaddrinfo: "1.centos.pool.ntp.org" invalid host address, ignored: 2 time(s)
    getaddrinfo: "2.centos.pool.ntp.org" invalid host address, ignored: 2 time(s)

  ---------------------- XNTPD End -------------------------

Yet from this host I get:

# host 0.centos.pool.ntp.org
0.centos.pool.ntp.org has address 208.53.158.34
0.centos.pool.ntp.org has address 69.50.219.51
0.centos.pool.ntp.org has address 108.76.168.145

So what is NTP doing here; why the errors for getaddrinfo?

Or is this nothing more than 2 out of some many times getaddrinfo failed, but worked just fine enough to count?