I got it working again by using IP addresses instead of
hostnames in ntp.conf (server entries). I still have no
ideas why that makes any differences. The nscd hosts cache
is disabled so there should be no cache. I still get the
'clnt_dg_create: out of memory' messages but that doesn't
seem to affect anything.
Thanks to
Sanjaya Srivastava <Sanjaya.Srivastava@Eng.Sun.COM>
mwang@tech.cicg.ml.com (Michael Wang)
for helping out.
Viet Hoang
vhoang@lucent.com
Original message
=> Date: Thu, 05 Nov 1998 10:43:02 -0700
=>
=> A machine name was mistakenly added to our NIS hosts map
=> with an IP address of 0.0.0.0. This somehow broke our
=> ntp setup. We took out the entry, rebuilt the maps and
=> pushed them. On 1 machine, we got xntpd to work again
=> by stop & restart nscd & xntpd. On another machine,
=> this would not fix it, here's the error message
=>
=> # ntptrace
=> localhost: stratum 16, offset 0.000093, synch distance 1.10953
=> 0.0.0.0: *Timeout*
=>
=> Does xntpd keep a cache of server addresses somewhere
=> that needs to be cleared out? The machines are E5000's
=> running Solaris 2.6 with the recommended patches.
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:12:52 CDT