All OnePasses started 2 minutes late last night

Not in a few months: WARNING TiVo clock running two minutes fast

When was your last service connection?

I just checked the time servers and all were within 2 seconds.

 

At a command prompt, you can see the clock drift on TiVo's servers with this command:

w32tm /monitor /computers:sjr1.tivo.com,sjr2.tivo.com,sjr3.tivo.com,time3.apple.com

There is very long thread with much more detail: Clock off in Roamio by 2 minutes

Rather than forcing a service connection, you could just run the internet diagnostics under Network Troubleshooting. It only takes a minute and sets the clock.

 

Advertisements

You know I see this from time to time. What I don’t get is why we see it. There really should not be drift of any kind really. So what is the permanent fix going to be if any? Come on TiVo you can do it right can’t you?

 

I never see it. True, it might happen and then correct before I do a recording.

 

Joe
I just checked the servers and sjr1 is off about 14 seconds.

 

Darn. I thought that was fixed.

 

At a command prompt? Where??

 

windows command prompt

win-R cmd

 

Wow. That's really cool. I've never used that program/external command before. Here's my output after I forced my Window machine to sync via the GUI.

sjr1.tivo.com[204.176.49.10:123]:
ICMP: 14ms delay
NTP: +8.9293023s offset from local clock
RefID: 'INIT' [0x54494E49]
Stratum: 0
sjr2.tivo.com[204.176.49.11:123]:
ICMP: 15ms delay
NTP: +0.3346304s offset from local clock
RefID: clock.isc.org [204.152.184.72]
Stratum: 3
sjr3.tivo.com[204.176.49.12:123]:
ICMP: 15ms delay
NTP: +0.1762562s offset from local clock
RefID: time3.apple.com [17.254.0.31]
Stratum: 2
time3.apple.com[17.254.0.31:123]:
ICMP: 19ms delay
NTP: +0.0029116s offset from local clock
RefID: 'GPSs' [0x73535047]
Stratum: 1

 

What we saw in the past is some of the NTP servers would drift up to a couple of minutes, so you had to monitor over time as over a couple of hours they would drift from a few seconds off to a few minutes off.

I don't know if that is what is going on now but if you have a suspect server you can use the following to monitor continuously

w32tm /stripchart /computer:sjr1.tivo.com

w32tm /stripchart /dataonly /computer:sjr1.tivo.com

Substitute sjr1 sjr2 sjr3 as appropriate.

First command displays a text-based graphic that illustrates how much off the server is from your machine. The second displays just the offsets with no text-based graphic

If it is the same problem as before, the offsets will just keep increasing over time until they reach a max and then reset to close to accurate time.

Not everyone would see that problem because at least one of the other servers would have accurate time and in certain scenarios it was random which one your TiVo decided to use when the 3 didn't agree, so you could reconnect to mothership over and over and some would get less accurate time and some would get more accurate time.

 

My recordings this morning started around 20 seconds too late.

 

The post from cwerdna showed 14 milliseconds, not 14 seconds.

 

That was packet transit time reference. The offset on sjr1 was 8.9seconds from a just synced local clock.

 

My bad, but 8.9 seconds still doesn't mean 2 minutes. Something else has to be going on here.

 

So I blacklisted sjr1, sjr2 and sjr3 on my router and was still able to do a successful service connection. I believe there should have been an error. Maybe TiVo is still using some other time servers.

 

I'm not the best source for help since I have never had a time problem.

 

Joe,

Is this relevant anymore (TiVo's ntp servers)? I thought they finally switched to using some of the standard Internet ntp servers.

This post shows them using the NIST ntp servers.

Clock off in Roamio by 2 minutes

Scott