FAQ Search Today's Posts Mark Forums Read
» Video Reviews

» Linux Archive

Linux-archive is a website aiming to archive linux email lists and to make them easily accessible for linux users/developers.


» Sponsor

» Partners

» Sponsor

Go Back   Linux Archive > Debian > Debian User

 
 
LinkBack Thread Tools
 
Old 09-20-2012, 02:56 PM
John Hasler
 
Default ntpd crashes.

Someone who Stefan failed to identify wrote:
> ...ntpd crashes on my server. Time jumps forward one hour every time
> this has happened.

I doubt ntpd is crashing. Most likely something else is jumping the
system clock and ntpd is behaving as designed and exiting when it sees a
one hour error. Shut down ntpd, set the clock correctly, and wait. I
suspect that eventually you'll see the clock jump even without ntpd
running.
--
John Hasler


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 87627868vb.fsf@thumper.dhh.gt.org">http://lists.debian.org/87627868vb.fsf@thumper.dhh.gt.org
 
Old 09-20-2012, 08:20 PM
Mauro
 
Default ntpd crashes.

On 20 September 2012 16:56, John Hasler <jhasler@newsguy.com> wrote:
> Someone who Stefan failed to identify wrote:
>> ...ntpd crashes on my server. Time jumps forward one hour every time
>> this has happened.
>
> I doubt ntpd is crashing. Most likely something else is jumping the
> system clock and ntpd is behaving as designed and exiting when it sees a
> one hour error. Shut down ntpd, set the clock correctly, and wait. I
> suspect that eventually you'll see the clock jump even without ntpd
> running.
> --
> John Hasler

Yes that is.
I have no cron jobs at all, I only have xen and heartbeat + pacemaker
on my two nodes and no processes that set the system clock.
The clock jumps ahead of one or two hours and I don't know why.
Perhaps some kernel bug?


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: http://lists.debian.org/CAE17a0UfLsG9MjRYstubqqeYwSacaybwjhuMg=G=K2aV4cWLn Q@mail.gmail.com
 
Old 09-21-2012, 10:05 AM
Mauro
 
Default ntpd crashes.

On 20 September 2012 22:20, Mauro <mrsanna1@gmail.com> wrote:
> On 20 September 2012 16:56, John Hasler <jhasler@newsguy.com> wrote:
>> Someone who Stefan failed to identify wrote:
>>> ...ntpd crashes on my server. Time jumps forward one hour every time
>>> this has happened.
>>
>> I doubt ntpd is crashing. Most likely something else is jumping the
>> system clock and ntpd is behaving as designed and exiting when it sees a
>> one hour error. Shut down ntpd, set the clock correctly, and wait. I
>> suspect that eventually you'll see the clock jump even without ntpd
>> running.
>> --
>> John Hasler
>
> Yes that is.
> I have no cron jobs at all, I only have xen and heartbeat + pacemaker
> on my two nodes and no processes that set the system clock.
> The clock jumps ahead of one or two hours and I don't know why.
> Perhaps some kernel bug?

It seems not ntp problem but a kernel bug:

http://my.opera.com/marcomarongiu/blog/2010/08/18/debugging-ntp-again-part-4-and-last


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: http://lists.debian.org/CAE17a0WcpqaoSK6ZJsT0MTmLHmPyqOH=tJeiv0bZyqx9Yu1Ff A@mail.gmail.com
 
Old 09-21-2012, 02:30 PM
Camaleón
 
Default ntpd crashes.

On Fri, 21 Sep 2012 12:05:01 +0200, Mauro wrote:

> On 20 September 2012 22:20, Mauro <mrsanna1@gmail.com> wrote:
>> On 20 September 2012 16:56, John Hasler <jhasler@newsguy.com> wrote:
>>> Someone who Stefan failed to identify wrote:
>>>> ...ntpd crashes on my server. Time jumps forward one hour every time
>>>> this has happened.
>>>
>>> I doubt ntpd is crashing. Most likely something else is jumping the
>>> system clock and ntpd is behaving as designed and exiting when it sees
>>> a one hour error. Shut down ntpd, set the clock correctly, and wait.
>>> I suspect that eventually you'll see the clock jump even without ntpd
>>> running.
>>
>> Yes that is.
>> I have no cron jobs at all, I only have xen and heartbeat + pacemaker
>> on my two nodes and no processes that set the system clock. The clock
>> jumps ahead of one or two hours and I don't know why. Perhaps some
>> kernel bug?

ntpd should avoid the clock going that forward and keep it disciplined.

> It seems not ntp problem but a kernel bug:
>
> http://my.opera.com/marcomarongiu/blog/2010/08/18/debugging-ntp-again-part-4-and-last

You can try the mentioned work-around and see if that works for you.

Anyway, if that's the case, you should experience the same with different
ntp daemons and not just with ntpd :-?

Greetings,

--
Camaleón


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: http://lists.debian.org/k3htlb$oni$3@ger.gmane.org
 
Old 09-22-2012, 01:51 PM
Camaleón
 
Default ntpd crashes.

El 2012-09-21 a las 22:13 +0200, Mauro escribió:

(resending to the list)

> On 21 September 2012 16:30, Camaleón <noelamac@gmail.com> wrote:
> > On Fri, 21 Sep 2012 12:05:01 +0200, Mauro wrote:
> >
> >> On 20 September 2012 22:20, Mauro <mrsanna1@gmail.com> wrote:
> >>> On 20 September 2012 16:56, John Hasler <jhasler@newsguy.com> wrote:
> >>>> Someone who Stefan failed to identify wrote:
> >>>>> ...ntpd crashes on my server. Time jumps forward one hour every time
> >>>>> this has happened.
> >>>>
> >>>> I doubt ntpd is crashing. Most likely something else is jumping the
> >>>> system clock and ntpd is behaving as designed and exiting when it sees
> >>>> a one hour error. Shut down ntpd, set the clock correctly, and wait.
> >>>> I suspect that eventually you'll see the clock jump even without ntpd
> >>>> running.
> >>>
> >>> Yes that is.
> >>> I have no cron jobs at all, I only have xen and heartbeat + pacemaker
> >>> on my two nodes and no processes that set the system clock. The clock
> >>> jumps ahead of one or two hours and I don't know why. Perhaps some
> >>> kernel bug?
> >
> > ntpd should avoid the clock going that forward and keep it disciplined.
> >
> >> It seems not ntp problem but a kernel bug:
> >>
> >> http://my.opera.com/marcomarongiu/blog/2010/08/18/debugging-ntp-again-part-4-and-last
> >
> > You can try the mentioned work-around and see if that works for you.
> >
> > Anyway, if that's the case, you should experience the same with different
> > ntp daemons and not just with ntpd :-?
>
> I've experienced the same problem also with openntp.

That makes more sense.

Anyway, no NTP daemon should crash because of skewed time; one thing is
that it refushes to sync (which can be fine, and should log this fact
so the admin can make the proper measures) but a different thing is
completely killing the service.

> Now I'm back with ntp using the workaround mentioned, hope it works.

Perfect, tell us how it went.

Greetings,

--
Camaleón


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120922135126.GB5812@stt008.linux.site">http://lists.debian.org/20120922135126.GB5812@stt008.linux.site
 
Old 09-23-2012, 09:08 AM
Rick Thomas
 
Default ntpd crashes.

On Sep 22, 2012, at 6:51 AM, Camaleón wrote:

Anyway, no NTP daemon should crash because of skewed time; one thing
is

that it refushes to sync (which can be fine, and should log this fact
so the admin can make the proper measures) but a different thing is
completely killing the service.



Hi Camaleón! A bit of NTP history:

That issue has been argued on the NTP developer mailing lists.
Crashing the daemon is Dave Mills' way of telling the admin that
something is badly broken here and needs to be fixed. Several
developers (myself included) disagreed with him at the time, but he
was adamant on the subject. So that's the way it is.


Rick

--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: D8B34071-ED17-429B-A6E9-0F602E37F502@pobox.com">http://lists.debian.org/D8B34071-ED17-429B-A6E9-0F602E37F502@pobox.com
 
Old 09-23-2012, 11:28 AM
Neal Murphy
 
Default ntpd crashes.

On Sunday, September 23, 2012 05:08:19 AM Rick Thomas wrote:
> On Sep 22, 2012, at 6:51 AM, Camaleón wrote:
> > Anyway, no NTP daemon should crash because of skewed time; one thing
> > is
> > that it refushes to sync (which can be fine, and should log this fact
> > so the admin can make the proper measures) but a different thing is
> > completely killing the service.
>
> Hi Camaleón! A bit of NTP history:
>
> That issue has been argued on the NTP developer mailing lists.
> Crashing the daemon is Dave Mills' way of telling the admin that
> something is badly broken here and needs to be fixed. Several
> developers (myself included) disagreed with him at the time, but he
> was adamant on the subject. So that's the way it is.

I must agree that that was a somewhat braindead decision. 'Failure by design'
is never an valid option.

If a problem is important enough to require an admin's attention, a daemon
should demand it via the console and the system logs. Nag repeatedly and
incessantly until the problem is corrected. Interrupt booting until the admin
has acknowledged the problem. But never fail silently.


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 201209230728.13002.neal.p.murphy@alum.wpi.edu">htt p://lists.debian.org/201209230728.13002.neal.p.murphy@alum.wpi.edu
 
Old 09-23-2012, 12:46 PM
Henrique de Moraes Holschuh
 
Default ntpd crashes.

On Sun, 23 Sep 2012, Neal Murphy wrote:
> I must agree that that was a somewhat braindead decision. 'Failure by design'
> is never an valid option.
>
> If a problem is important enough to require an admin's attention, a daemon
> should demand it via the console and the system logs. Nag repeatedly and

Do that and you will DoS the system logs. You log it *ONCE* using the
appropriate facility and severity.

--
"One disk to rule them all, One disk to find them. One disk to bring
them all and in the darkness grind them. In the Land of Redmond
where the shadows lie." -- The Silicon Valley Tarot
Henrique Holschuh


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120923124633.GA3939@khazad-dum.debian.net">http://lists.debian.org/20120923124633.GA3939@khazad-dum.debian.net
 
Old 09-23-2012, 01:13 PM
"David L. Craig"
 
Default ntpd crashes.

On 12Sep23:0208-0700, Rick Thomas wrote:
>
> On Sep 22, 2012, at 6:51 AM, Camaleón wrote:
>
> >Anyway, no NTP daemon should crash because of skewed time;
> >one thing is that it refushes to sync (which can be fine,
> >and should log this fact so the admin can make the proper
> >measures) but a different thing is completely killing the
> >service.
>
> That issue has been argued on the NTP developer mailing lists.
> Crashing the daemon is Dave Mills' way of telling the admin
> that something is badly broken here and needs to be fixed.
> Several developers (myself included) disagreed with him at the
> time, but he was adamant on the subject. So that's the way it
> is.

There must be at least one patch available and maintained
that can be applied to modify the official distribution's
behavior if multiple developers disagree with this
behavior. Where should one look?
--
May the LORD God bless you exceedingly abundantly!

Dave_Craig________________________________________ ______
"So the universe is not quite as you thought it was.
You'd better rearrange your beliefs, then.
Because you certainly can't rearrange the universe."
__--from_Nightfall_by_Asimov/Silverberg_________________


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120923131356.GA31628@dlc-dt.home">http://lists.debian.org/20120923131356.GA31628@dlc-dt.home
 
Old 09-23-2012, 03:36 PM
Camaleón
 
Default ntpd crashes.

On Sun, 23 Sep 2012 02:08:19 -0700, Rick Thomas wrote:

> On Sep 22, 2012, at 6:51 AM, Camaleón wrote:
>
>> Anyway, no NTP daemon should crash because of skewed time; one thing is
>> that it refushes to sync (which can be fine, and should log this fact
>> so the admin can make the proper measures) but a different thing is
>> completely killing the service.
>
>
> Hi Camaleón! A bit of NTP history:
>
> That issue has been argued on the NTP developer mailing lists. Crashing
> the daemon is Dave Mills' way of telling the admin that something is
> badly broken here and needs to be fixed. Several developers (myself
> included) disagreed with him at the time, but he was adamant on the
> subject. So that's the way it is.

Wow... I didn't know about the reason "in behind", thanks. Needless to
say I also find it a bit radical measure for dealing with a problem.
Should we also evaluate a kernel error in that way, no system would be up
for more than... a week? :-)

Either way, ntpd provides some options to conciliate with severely skewed
time which can help to prevent (or mitigate) the daemon to crash.

Greetings,

--
Camaleón


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: http://lists.debian.org/k3na93$lsj$9@ger.gmane.org
 

Thread Tools




All times are GMT. The time now is 11:03 AM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.
Copyright ©2007 - 2008, www.linux-archive.org