NTP synchronization issue

[derek wilson]

Hi

Got a new datacentre build  2 rtr's and 2 sw's go external for timing -everything else infrastructure wise looks to those rtr's and sw's for timing

All units are solid apart from 2 small 3560's that keep dropping sync -

Sep 21 15:53:03.966: NTP message sent to x.x.x.8, from interface 'Vlan101' (x.x.x.7).
Sep 21 15:53:03.968: NTP message received from x.x.x.8 on interface 'Vlan101' (x.x.x.7).
Sep 21 15:53:03.968: NTP Core(DEBUG): ntp_receive: message received
Sep 21 15:53:03.968: NTP Core(DEBUG): ntp_receive: peer is 0x3D163DE0, next action is 1.
Sep 21 15:53:03.969: NTP Core (NOTICE): Clock synchronization lost.
.Sep 21 15:53:03.969: NTP Core (INFO): x.x.x.8 961A 8A sys_peer
.Sep 21 15:53:03.969: NTP Core (NOTICE): Clock is synchronized.
Sep 21 15:53:07.969: NTP message sent to x.x.x.2, from interface 'Vlan101' (x.x.x.7).
Sep 21 15:53:07.971: NTP message received from x.x.x.2 on interface 'Vlan101' (x.x.x.7).
Sep 21 15:53:07.971: NTP Core(DEBUG): ntp_receive: message received

Anyone come across this before?  Strange as you can see on the above log (with debug on) that the sw recieves a ntp message then fails sync at the same point.

thanks

[iportuga]

Hi,

You need collect additional data to perform a proper troubleshooting:

• Check the software version and look for known issues;
• Collect additional debug and command outputs (such as 'show ntp association detail');
• Collect a network trace of the NTP traffic is also useful (at network element and/or at NTP peer/server);

Reference: Network Time Protocol (NTP) Issues Troubleshooting and Debugging Guide