Clock widget wrong on C40

[Darren McKinnon]

I have a client who has an Sx20 and C40 on the same LAN pointed to the same NTP server, but only the SX20 is displaying the correct time.. the C40 is off by about 41 minutes.. I've tried other time servers, and upgrading the software (currently 7.0 2) but nothing will cure this.. i have disabled the NTP settings altogether, but my client would like for it to be rectified if possible.

Any ideas what could be happening?

Darren

[Wayne DeNardi]

Hi Darren,

The usual cause of this from what I have seen in my environments is that the device not actually talking to the NTP server, even though it is configured - possibly a firewall or other connectivity issue between that endpoint and the NTP server.  If you check the web interface on the codec Diagnostics - > Troubleshooting - you'll often see a NTP status error.

Wayne
--
Please remember to rate responses and to mark your question as answered if appropriate.

[Darren McKinnon]

Wayne, I have checked the troubleshooting page, and nothing in there mentions NTP.  Did you mean to look within a specific log file?  Can you advise which one if yes.

[Wayne DeNardi]

Might be a TC7.1.x thing.  It's there on mine.  I've seen a couple recently have NTP problems and they showed there.

[Darren McKinnon]

I have upgraded to TC7.1.1, and now Troubleshooting does show that the NTP is not configured correctly, but I do not see any difference from the SX20 on the same network. 

I have opened a ticket with Cisco TAC.

[Wayne DeNardi]

Give TC7.1.1 a few minutes - I've seen many of mine report NTP issues on the Troubleshooting page upon first boot that then go away after a little while.  Check back on your codec after 5 or so minutes and re-run the diagnostics - you might see the error disappear if it is working (or in your case, if you've got an actual problem, maybe not).

Was there any difference setting your NTP Mode to "Manual" rather than "Auto"?

Do you have an AD server, or other time source on your network that you can point the endpoint to rather than the external Tandberg one?

Wayne
--
Please remember to rate responses and to mark your question as answered if appropriate.

[Bharadwaj GK]

SX10 standalone. 

On Static IP's 

Can ping all 3 NTP servers from the VC unit. 

DHCP off.

However, status is as below

xstat //ntp
*s NetworkServices NTP CurrentAddress: "45.125.255.53"
*s NetworkServices NTP Server 1 Address: "161.202.144.189"
*s NetworkServices NTP Server 2 Address: "139.59.15.228"
*s NetworkServices NTP Server 3 Address: "0.ntp.tandberg.com"
*s NetworkServices NTP Status: Discarded
** end

Pls advice. 

Any help here will be appreciated guys :) 

[juriss]

We have also seen some stubborn endpoints that do not seem to want to to NTP sync...

We typically use the IP Gateway address as the NTP source

My understanding is that any Cisco IOS device (Router or Switch) that is "in NTP sync" can be used as an NTP source.  This arrangement keeps all the NTP chatter on the LAN rather than out on the Internet or to an NTP server in your data center, this also avoids the DNS issue..   we put the IPv4 IP Gateway address in as the NTP server and it works like a charm 

In our case we have Layer 2 Cisco 3750 switches in the comm closet and we know that our network team always sets these Layer 2 switches up for NTP sync.

Hope this helps

SCJ

[Steve Kapinos]

Sometimes it's the simple stuff...  is the NTP a remote LAN?  Are you sure the C40 has it's gateway and subnet properly defined?  IE: the c40 talks to local LANs fine, but not certain remote LANs?

It's gotta be NTP config settings, network settings in endpoint, or connectivity between.  Check for simple transpositions, etc

[Patrick Sparkman]

If using a static IP on the codec, people sometimes forget to put in static DNS addresses as well.  I should know, happened to us a few years ago.

[Darren McKinnon]

Patrick, I think you nailed it.  My C40 did not have anything in the DNS settings.  I have just updated them, and rebooted the endpoint.  Is there a way I can verify the time remotely?

Darren

[Patrick Sparkman]

I'm running TC6.3.1, there is an SSH command to check ntp.

systemtools ntp

[Darren McKinnon]

after updating the DNS settings, and rebooting the codec, the time the codec is reporting when using the systemtools ntp is 9 days old, (current time is 11:58 PM ADT, April 17h, codec says 11:53 ADT, April 9th.

I started extended logging on the C40, and here is something I found in the eventlog/all

Apr  9 23:53:39.317 ppc eventlog[3032]: ntp: recv_response: poll() failed or timed out.
Apr  9 23:53:39.318 ppc eventlog[3032]: ntp: Failed to recv READSTAT response from NTP daemon.
Apr  9 23:53:39.318 ppc eventlog[3032]: ntp: READSTAT failed.

[Martin Koch]

What is your NTP setting?

If its auto, NTP server settings might be taken from DHCP, so if they are wrong it might fail anyhow.

Besides that, check that you can properly resolve the host names and that the

firewall is open. A tcpdump can also help you to see what going on or not.

What do (loged in via ssh as admin)

xconf // ntp
xstat // ntp
systemtools ntp

say?

[Darren McKinnon]

Here are the settings from the C40.  I compared these to the SX20 also on the same network, and they match exactly, except for the current date and time - it is 8 hours and approx 5 minutes old.

xconfig // ntp
*c xConfiguration Network 1 QoS Diffserv NTP: 0
*c xConfiguration NetworkServices NTP Mode: Auto
*c xConfiguration NetworkServices NTP Address: "1.ntp.tandberg.com"
** end

OK
XSTAT // NTP
*s NetworkServices NTP Address: "1.ntp.tandberg.com"
*s NetworkServices NTP CurrentAddress: "127.127.1.0"
*s NetworkServices NTP Status: Synced
** end

OK
SYSTEMTOOLS NTP
     remote           refid      st t when poll reach   delay   offset  jitter
==============================================================================
*127.127.1.0     LOCL            10 l   14   16  377    0.000    0.000   0.000

Current time in UTC is: Mon Apr 14 23:37:00 2014
Current time in ADT is: Mon Apr 14 20:37:00 2014

OK