Showing results for 
Search instead for 
Did you mean: 

Nexus NTP offset wrong

  This is a bit different than the other Nexus NTP issues that came up when searching the Community forums. Here is the scenario:

I'm in the Eastern timezone (EST). Time is correct on the Windows DC that we are using as a time server. Config on the Nexus is thus:

clock timezone EST -5 0

clock summer-time EDT 2 Sun Mar 02:00 1 Sun Nov 02:00 60

However, when I perform show clock I see this

TEST-Core(config)# sh clock

20:55:33.483 EDT Thu Sep 05 2013

... when the time is actually 2:55pm EDT.  I have to set clock timezone EDT -11 0 (which is wrong) to get the correct time.  Anyone seen this or have an idea what's happened? Thanks...

Marvin Rhoads
VIP Community Legend

Mine are working OK. What does "show ntp peer-status" show?

Total peers : 2

* - selected for sync, + -  peer mode(active),

- - peer mode(passive), = - polled in client mode

    remote               local                 st   poll   reach delay   vrf


=            2   16     377   0.00067 default

=            1   16     377   0.00073 default

Marvin Rhoads
VIP Community Legend

So your NTP servers arent syncing. Otherwise we would expect to see one of them to have the asterisk (*) showing that it is selected for sync. Once that is fixed, you can recheck the status.

The NTP servers are syncing with the time source. Perhaps I'm not understanding what you mean.

Marvin Rhoads
VIP Community Legend

The NTP servers are synced with the time source but not with the Nexus. When a switch is synchronized with an NTP server it indicates so (as noted in the output you provided) by putting an asterisk in front of the server it has selected for sync.

I look for the asterisk when checking ntp server synchronization status from a switch or router:

     show ntp peer-status (NX-OS)

     show ntp association (IOS)

  So I guess the question is, how do I get it to synchronize?  I've been configuring NTP on IOS for years, never had this much difficulty with it...

Marvin Rhoads
VIP Community Legend

I have seen issues with the ntp serversneedsing to be specified to be reached via the mgmt VRF. The default is to use the default VRF. When you have a dedicated management port (and/or VRF) you sometimes need to specify its use explicitly - e.g.:

     ntp server use-vrf mgmtVRF described here.