Clock time if off despite using "set automatically"

I have my clock set to use “Public Time Server (pool.ntp.org)” but it’s actually off by 10 minutes (and I think it’s getting off by more and more each day). Any idea what’s wrong and how to fix this?

I had previously had it set manually, but it was getting off by more and more every day, so I thought using automatic time set would fix the issue.

How do I diagnose/fix this?

Set it manually.

I think you didn’t read my whole post. :slight_smile:

I had previously had it set manually, but it was getting off by more and more every day, so I thought using automatic time set would fix the issue.

Did you use YaST to configure this? YaST normaly can check if the server works, did that succeed?
Show the contents of the config file:

grep -v '^#' ntp.conf

May indicate a MB battery going bad.

Why should that matter if it’s setting it via the network?

Here’s the contents (and no, I didn’t know YAST could do this. I just “right-clicked” on the clock):

server 127.127.1.0 # local clock (LCL)
fudge 127.127.1.0 stratum 10 # LCL is unsynchronized

driftfile /var/lib/ntp/drift/ntp.drift # path for drift file

logfile /var/log/ntp # alternate log file

keys /etc/ntp.keys # path for keys file
trustedkey 1 # define trusted keys
requestkey 1 # key (7) for accessing server variables

BUt we are not certain it is if it was the time should be correct not 10 or so minuets off.

On 2011-02-23 16:06, 6tr6tr wrote:
>
> stamostolias;2294175 Wrote:
>> Set it manually.
>
> I think you didn’t read my whole post. :slight_smile:
>
>> I had previously had it set manually, but it was getting off by more
>> and more every day, so I thought using automatic time set would fix the
>> issue.

If the clock is off by too much, the ntp daemon will refuse to set up the
clock - thus, again, do it manually.

After that is done, explain your daily procedure.

Do you boot each day, and just after boot the clock is bad? Or is your
machine continuously on?

Plus. On the ntp daemon, configure several remote servers - in the ntp.conf
file you posted there is none. With that configuration, it will not work.


Cheers / Saludos,

Carlos E. R.
(from 11.2 x86_64 “Emerald” at Telcontar)

The above people are corect:

  1. you have no NTP server configured.
  2. when the the time is off for such a big distance, NTP will not redress that in one big bang.
  3. when you have a time difference of about 10 mins a day, I also guess your CMOS battery is near dead and should be replaced (for more reasons then only the time shift).

And you placed the computer output of the command I gave you between QUOTE tags. Please do this next time between CODE tags to keep the computer lay-out: Posting in Code Tags - A Guide.

On Wednesday 23 Feb 2011 15:06, 6tr6tr scribbled:

>
> I have my clock set to use “Public Time Server (pool.ntp.org)” but it’s
> actually off by 10 minutes (and I think it’s getting off by more and
> more each day). Any idea what’s wrong and how to fix this?
>
> I had previously had it set manually, but it was getting off by more
> and more every day, so I thought using automatic time set would fix the
> issue.
>
> How do I diagnose/fix this?
>
>

In the NTP configuration in YAST, what have you selected for “start NTP
daemon”? I’ve set it as “now and on boot”.
When I’ve had this happen to me -a few years ago now - I’ve just gone
through the configuration again and the time gets reset on clicking “OK”.


Graham Davis, Bracknell, Berks. E-mail: change boy to man

Thanks!

The issue is that for some reason in KDE, there’s about 5 different places where you configure what appear to be the same things but are not and are all related to date/time. I have no idea why they’re not all aggregated into one app.

So, using YAST, I added some ntp servers including “0.us.pool.ntp.org” (and 1., etc). Then I chose “synchronize now” and it fixed the time!

Thanks everyone!

But I still ythinks that 10 mins per day (when not corrected NTP) is something that shouuld worry you aboyt the status of the CMOS battery.

About those different applications/places to do somethiing with date/time, I do not have the ultimate answer because I do not know all of them you found. But you should be aware of the fact that setting the NTP configuration requires root actions because it is a system thingy configured in system files. Thus, whichever tool you use, as long as it does not ask for the root password, it does not configure NTP. Also you should not forget that apart from keeping the system clock synchronized to UTC (a system task where NTP helps), individual users can set their environment to live in a certain time zone. And that is where the desktop software (KDE, Gnome, …) can offer assistance. But that has nothing to do witth the system time, but with the way the time is shown to the end-user (time zone, daylight saving time). This is often difficult to understand by people used to systems not coping for time management for indivudual end-users (and that is not only Windows, but even IBM 360 systems suffer from that).

On 2011-02-26 10:36, hcvv wrote:
>
> But I still ythinks that 10 mins per day (when not corrected NTP) is
> something that shouuld worry you aboyt the status of the CMOS battery.

Not necessarily.

The battery is only used while the computer is OFF, and the error appears
on boot. But it can also be that the /etc/adjtime file has wrong values.
Notice that a bad battery should produce an error in the BIOS boot screen.

Other causes of error are simply that the system clock (it is not related
at all to the battery) runs fast or slow, because of some hardware or
software problem, or that it has been badly or not disciplined by NTP yet.


Cheers / Saludos,

Carlos E. R.
(from 11.2 x86_64 “Emerald” at Telcontar)

On 26/02/11 15:03, Carlos E. R. wrote:
> On 2011-02-26 10:36, hcvv wrote:
>>
>> But I still ythinks that 10 mins per day (when not corrected NTP) is
>> something that shouuld worry you aboyt the status of the CMOS battery.
>
> Not necessarily.
>
> The battery is only used while the computer is OFF, and the error appears
> on boot. But it can also be that the /etc/adjtime file has wrong values.
> Notice that a bad battery should produce an error in the BIOS boot screen.
>

My experience differs: the first indication of a “low” battery is that
the clock starts losing time. The next stage is, sometimes, that the
system will not boot until the 5vSBY has been present for a short while
and then the system may self start - the CMOS settings are retained.
Finally, the voltage drops to the point where the system will boot
reliably but the CMOS settings are lost and the BIOS reports an error at
boot time.

> Other causes of error are simply that the system clock (it is not related
> at all to the battery) runs fast or slow, because of some hardware or
> software problem, or that it has been badly or not disciplined by NTP yet.
>


PeeGee

MSI m/b 870-C45, AMD Athlon II X3 445, 4GB, openSUSE 11.3/11.2 x86_64
dual boot + XP Home in VBox
Asus m/b M2NPV-VM, AMD 64X2 3800+, 2GB, openSUSE 11.3 x86_64/XP Home
dual boot
Acer Aspire 1350, AMD (M)XP2400+, 768MB, openSUSE 11.2/XP Home dual boot
Asus eeePC 4G (701), Celeron M353, 2GB, openSUSE 11.3 on SSD

PeeGee, could that be the cause of this? Getting errors from hard drive!

On 2011-02-26 18:29, PeeGee wrote:
> On 26/02/11 15:03, Carlos E. R. wrote:

> My experience differs: the first indication of a “low” battery is that
> the clock starts losing time. The next stage is, sometimes, that the
> system will not boot until the 5vSBY has been present for a short while
> and then the system may self start - the CMOS settings are retained.
> Finally, the voltage drops to the point where the system will boot
> reliably but the CMOS settings are lost and the BIOS reports an error at
> boot time.

In the cases I have seen, the BIOS, or the system, warned. In some cases,
bios config was set to defaults and clock reset to 1980 or so. I have never
seen a slow clock. And, if you are using ntp, you would not notice.

The clock technology is the same as with you hand watch: they work fine
till one day they stop completely.


Cheers / Saludos,

Carlos E. R.
(from 11.2 x86_64 “Emerald” at Telcontar)

On 2011-03-01 16:36, 6tr6tr wrote:
>
> PeeGee;2295635 Wrote:

> PeeGee, could that be the cause of this? ‘Getting errors from hard
> drive!’ (http://tinyurl.com/4lujz3l)

Please explain your theory. How can a bad clock cause disk errors?


Cheers / Saludos,

Carlos E. R.
(from 11.2 x86_64 “Emerald” at Telcontar)