[Planetlab-users] NTP daemon

[Alex Sherman]

Speaking of NTP I consistently see some machines with several minutes to
an hour drift compared to the rest of the machines that are typically in
sync: 

planetlab2.inf.ethz.ch (-10 minutes) 
planetlab3.inf.ethz.ch (-10 minutes) 
planetlab2.ukc.ac.uk (+ 55 minutes) 

Are these talking to an NTP server ? 

thanks 
-Alex


On Wed, 2008-01-30 at 01:34 -0500, Neil Spring wrote:
> Markus,
> 
> NTP is already running on planetlab... believe it or not.
> 
> You may be able to run some of the ntp tools (e.g., ntptrace) on  
> planetlab to figure out what the running daemon thinks.
> 
> (Even if ntp weren't already running, you would need permission to  
> bind the privileged port 123.  I don't *think* there's a way to run an  
> NTP server on a different port, but you might look into it.)
> 
> -neil
> 
> On Jan 29, 2008, at 5:30 PM, Markus Gnauck wrote:
> 
> > Hello PlanetLab Users!
> >
> > I installed ntp on some of my project's nodes. Unfortunately when  
> > trying to run the ntp service it seems to crash immediately upon  
> > startup.
> >
> > $ sudo yum install ntp
> > ..works as expected..
> >
> > $ sudo /etc/init.d/ntpd start
> > Starting ntpd:                                             [  OK  ]
> >
> > $ /etc/init.d/ntpd status
> > ntpd dead but pid file exists
> >
> > I tried to search for the problem, with some sources suggesting to  
> > delete /var/run/ntpd.pid and restarting ntpd. As anticipated that  
> > did not help much. The results are still as depicted above. (To my  
> > knowledge the configuration file ntpd.conf seems fine though.)
> >
> > When tracing the problem my tutor came up with the following output  
> > which strongly seems like a permission issue:
> >
> > Jan 27 18:42:54 chronos ntpd[31596]: ntpd 4.2.0a at 1.1196-r Fri May 12
> > 09:51:35 EDT 2006 (1)
> > Jan 27 18:42:54 chronos ntpd[31596]: Cannot set RLIMIT_MEMLOCK:
> > Operation not permitted
> > Jan 27 18:42:54 chronos ntpd[31596]: mlockall(): Cannot allocate  
> > memory
> > Jan 27 18:42:54 chronos ntpd[31596]: precision = 1.000 usec
> > Jan 27 18:42:54 chronos ntpd[31596]: bind() fd 4, family 2, port 123,
> > addr 0.0.0.0, in_classd=0 flags=8 fails: Permission denied
> > Jan 27 18:42:54 chronos ntpd[31596]: bind() fd 4, family 2, port 123,
> > addr 127.0.0.1, in_classd=0 flags=0 fails: Permission denied
> > Jan 27 18:42:54 chronos ntpd[31596]: bind() fd 4, family 2, port 123,
> > addr 134.34.246.5, in_classd=0 flags=8 fails: Permission denied
> > Jan 27 18:42:54 chronos ntpd[31596]: bind() fd 4, family 2, port 123,
> > addr 134.34.246.5, in_classd=0 flags=8 fails: Permission denied
> > Jan 27 18:42:54 chronos ntpd[31596]: bind() fd 4, family 2, port 123,
> > addr 10.41.221.1, in_classd=0 flags=8 fails: Permission denied
> > Jan 27 18:42:54 chronos ntpd[31596]: kernel time sync status 0040
> > Jan 27 18:42:54 chronos ntpd[31596]: cap_set_proc() failed to drop  
> > root
> > privileges: Operation not permitted
> >
> > The problem appeared on all the nodes that I've considered. I'm kind  
> > of at a loss right now. Do you have any suggestions, pointers or  
> > help to solve my problem?!
> > Please let me know if I better direct the question to a different  
> > source of support for issues regarding PlanetLab.
> >
> > Thanks in advance! Cheers,
> > -Markus
> >
> > _______________________________________________
> > Users mailing list: Users at lists.planet-lab.org
> > https://lists.planet-lab.org/mailman/listinfo/users
> 
> _______________________________________________
> Users mailing list: Users at lists.planet-lab.org
> https://lists.planet-lab.org/mailman/listinfo/users