[ntp:questions] NTP Server not responding to clients

James jp1046 at stl.rural.usda.gov
Thu Aug 7 17:10:05 UTC 2003


Apologies for the verbosity, but it's sort of required for this one, I
think.

I've got an NTP server that is not responding to client requests for
timesync. Here's the scoop:

servermachine is the linux NTP server.
clientmachine is a linux NTP client.
There are miscellaneous other NTP servers/peers/clients around.

Here's the machine vitals:
-----
[root at servermachine log]# uname -a
Linux servermachine 2.4.2-2smp #1 SMP Sun Apr 8 20:21:34 EDT 2001 i686
unknown
 
[root at servermachine log]# rpm -q ntp
ntp-4.0.99k-15
-----

The server is configured to get time from some stratum 1s and peer to
some strat 2s:
-----
[root at servermachine log]# cat /etc/ntp.conf
server  127.127.1.0     
fudge   127.127.1.0 stratum 10
driftfile /etc/ntp/drift
multicastclient         
broadcastdelay  0.008
authenticate no
server  foo.stratum1.com
server  bar.stratum1.com
peer    foo.stratum2.com
peer    bar.stratum2.com
peer    baz.stratum2.com
-----

There are no firewall rules blocking anything at this time:
-----
[root at servermachine log]# iptables -L -v
Chain INPUT (policy ACCEPT 1253 packets, 111128 bytes)
 pkts bytes target     prot opt in     out     source              
destination
  
Chain FORWARD (policy ACCEPT 0 packets, 0 bytes)
 pkts bytes target     prot opt in     out     source              
destination
  
Chain OUTPUT (policy ACCEPT 532 packets, 64847 bytes)
 pkts bytes target     prot opt in     out     source              
destination
 
[root at servermachine log]# /etc/init.d/ntpd start
Starting ntpd:                                             [  OK  ]
-----

As you can see here, this server is able to reach all of its servers
and its peers just fine:
-----
[root at servermachine log]# ntpq -p
     remote           refid      st t when poll reach   delay   offset
 jitter
==============================================================================
 LOCAL(0)        LOCAL(0)        10 l    4   64  377    0.000    0.000
  0.000
 foo.stratum1.co 0.0.0.0          1 u   16   64  377    0.000    0.000
4000.00
 bar.stratum1.co 0.0.0.0          1 u   60   64  377    0.000    0.000
4000.00
 foo.stratum2.co 0.0.0.0          2 u    1   64  377   71.070   -1.707
  0.000
 bar.stratum2.co 0.0.0.0          2 u   28   64  377    0.000    0.000
4000.00
 baz.stratum2.co 0.0.0.0          2 u    -   64    0    0.000    0.000
4000.00
-----

*THE PROBLEM*
Even though servermachine is able to sync with it's servers OK, none
of our network clients are able to sync with servermachine. As you can
see from the tcpdump below, the client NTP packets are arriving OK,
but servermachine just IGNORES them completely. No response is sent to
the client. No record is left in /var/log/messages.

NOTE: There are two tcpdumps below:
The first is a sample of normal traffic. You can see packets heading
out to the stratum1/stratum2 servers and you can see packets coming in
from miscellaneous clients.
The second tcpdump shows what it looks like when my linux machine,
clientmachine, runs a 'ntpdate servermachine' command. (The output on
the client end is "no server suitable for synchronization found".

[root at servermachine log]# tcpdump port 123
Kernel filter, protocol ALL, TURBO mode (575 frames), datagram packet
socket
tcpdump: listening on all devices
11:37:24.286374 eth0 < foo.stratum2.com.ntp > servermachine.com.ntp:
v3 sym_act strat 2 poll 10 prec -18 (DF)
11:37:32.466764 eth0 > servermachine.com.ntp > bar.stratum2.com.ntp:
v4 sym_act strat 0 poll 6 prec -17 (DF)
11:37:41.967216 eth0 < client1.1690 > servermachine.com.ntp: v2 client
strat 0 poll 11 prec 0
11:37:42.467240 eth0 > servermachine.com.ntp > foo.stratum2.com.ntp:
v4 sym_act strat 0 poll 6 prec -17 (DF)
11:37:45.787398 eth0 < client2.com.ntp > servermachine.com.ntp: v3
client strat 3 poll 10 prec -15 (DF)
11:37:46.967454 eth0 < client1.1690 > servermachine.com.ntp: v2 client
strat 0 poll 11 prec 0
11:37:47.467478 eth0 > servermachine.com.ntp > foo.stratum1.com: v4
client strat 0 poll 6 prec -17 (DF)
11:37:47.517480 eth0 < foo.stratum1.com > servermachine.com.ntp: v4
server strat 1 poll 6 prec -17 (DF) [tos 0x10]
11:37:49.657582 eth0 < client3.com.ntp > servermachine.com.ntp: v3
client strat 0 poll 6 prec -17 (DF)
11:37:51.967692 eth0 < client1.1690 > servermachine.com.ntp: v2 client
strat 0 poll 11 prec 0
11:37:56.087888 eth0 < client4.com.ntp > servermachine.com.ntp: v3
client strat 3 poll 10 prec -18 (DF)
  
11 packets received by filter
 
[root at appdoc2-stl log]# tcpdump port 123
Kernel filter, protocol ALL, TURBO mode (575 frames), datagram packet
socket
tcpdump: listening on all devices
11:49:41.722496 eth0 < clientmachine.ntp > servermachine.com.ntp: v4
client strat 0 poll 4 prec -6 (DF)
11:49:42.722546 eth0 < clientmachine.ntp > servermachine.com.ntp: v4
client strat 0 poll 4 prec -6 (DF)
11:49:43.722596 eth0 < clientmachine.ntp > servermachine.com.ntp: v4
client strat 0 poll 4 prec -6 (DF)
11:49:44.722646 eth0 < clientmachine.ntp > servermachine.com.ntp: v4
client strat 0 poll 4 prec -6 (DF)
  
4 packets received by filter
 
[root at servermachine log]#



More information about the questions mailing list