ntpd problems since upgrading to 5.3
Christian Hiris
4711 at chello.at
Mon Jan 17 18:21:13 PST 2005
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Tuesday 18 January 2005 01:09, John wrote:
> On Mon, Jan 17, 2005 at 12:49:00PM -0600, John wrote:
> > On Mon, Jan 17, 2005 at 12:22:48PM -0600, John wrote:
> > > On Tue, Jan 11, 2005 at 11:23:28PM +0900, Rob wrote:
> > > > Ian Moore wrote:
> > > > > Hi,
> > > > > Ever since I upgraded from 5.2.1-RELEASE to 5.3-RELEASE, I've been
> > > > > getting the following error on boot:
> > > > > ntpd[380]: bind() fd 7, family 28, port 123, addr fe80:1
> > > > >
> > > > > ::204:61ff:fe46:be89, in6_is_addr_multicast=0 flags=0 fails: Can't
> > > > > :: assign
> > > > >
> > > > > requested address
> > > > >
> > > > > ntpd seems to be working from what I can see in it's log file, but
> > > > > I can't do anything with ntpq to check it.
> > > > > Wether I run it as my normal user or as root, running ntpq -p
> > > > > always gives: ntpq: write to localhost.foo.com failed: Permission
> > > > > denied
Try to add "disable auth" to your ntp.conf.
> > > > I had once a problem with ntpd, when also running named. Some
> > > > hostname resolution failed, because the servers were started in the
> > > > wrong order. Are you also running named?
> > > >
> > > > > Here is my ntpd entries in rc.conf:
> > > > > ntpd_enable="YES" # Run ntpd Network Time Protocol
> > > > > (or NO). ntpd_program="/usr/sbin/ntpd" # path to ntpd, if you
> > > > > want a different one. ntpd_flags="-c /etc/ntp.conf -p
> > > > > /var/run/ntpd.pid"
> > > >
> > > > I use:
> > > > ntpd_enable="YES"
> > > > ntpd_flags="-g"
> > > >
> > > > > and the contents of ntp.conf:
> > > > > server 210.48.130.204
> > > > > server augean.eleceng.adelaide.edu.au
> > > > > driftfile /var/db/ntpd.drift
> > > > > logfile /var/log/ntpd
> > > >
> > > > And here I use:
> > > > driftfile /var/db/ntpd.drift
> > > > pidfile /var/run/ntpd.pid
> > > > server nr1.time.server
> > > > server nr2.time.server
> > > > server nr3.time.server
> > >
> > > OK - this is interesting!
> > >
> > > I have identical ntp.conf files on my 5.2.1 system and my 5.3-STABLE
> > > system. Guess what? The 5.2.1 system works, and the 5.3-STABLE system
> > > doesn't. Not only that, but the clock on my 5.3-STABLE system is
> > > RACING. It is going at almost twice as fast as real time.
> > >
> > > Here's the ntp.conf file:
> > > # stratum 3 time server
> > > server 192.168.1.1
> > >
> > > driftfile /var/db/ntp.drift
> > >
> > > In both cases, name resolution is working. On the 5.2.1 system, ntpdc
> > > shows:
> > > ntpdc> peers
> > > remote local st poll reach delay offset disp
> > > =======================================================================
> > > *dexter.starfire 192.168.1.52 3 64 377 0.00073 0.060184 0.00093
> > > ntpdc>
> > >
> > > On the 5.3-STABLE system, it ntpdc shows:
> > > ntpdc> peers
> > > remote local st poll reach delay offset disp
> > > =======================================================================
> > > =dexter.starfire 192.168.1.53 16 64 0 0.00000 0.000000 0.00000
> > > ntpdc>
> > >
> > > This shows that DNS is working fine, as the remote name is being
> > > correctly resolved. (I know I'm showing some of my IP numbers, but
> > > it's all NAT).
> > >
> > > I'm afraid something is broke!
> > >
> > > Oh, and ntpdate works on the 5.3 system just fine (when ntpd isn't
> > > running, of course).
> > >
> > > The system that is running 5.3-STABLE was a good time keeper before
> > > this update (4.9-STABLE).
> >
> > OK. An update.
> >
> > I ran
> > "ntpdate 192.168.1.1 ; ntpdate 192.168.1.1 ; ntpdate 192.168.1.1" and
> > suddenly, I'm keeping time MUCH better!
> >
> > My current theory is that whatever is going wrong with adjkerntz,
> > it messed up the kernel time keeping adjustment, and when I ran ntpdate
> > close enough together that it was able to use adjtime rather than
> > stepping the time, that helped things out greatly.
> >
> > ntpd still doesn't work, but my system is keeping time much better!
> > MUCH better!
>
> Stranger and stranger.
>
> Well, since ntp kept RUNNING, I neglected to check the logs. Shame on me.
>
> This is what goes into the log:
> Jan 17 18:04:29 pearl ntpd[838]: ntpd 4.2.0-a Sun Jan 9 10:58:59 CST 2005
> (1) Jan 17 18:04:29 pearl ntpd[838]: bind() fd 7, family 2, port 123, addr
> 0.0.0.0,in_classd=0 flags=8 fails: Address already in use
>
> HOWEVER, when I do a netstat -na | grep "\.12" before running it, there
> is no matches.
> After running it (and getting the error, but it stays running,
> and non-functional), I get:
> udp4 0 0 192.168.1.53.123 *.*
> udp6 0 0 fe80:5::206:25ff.123 *.*
> udp6 0 0 fe80:4::1.123 *.*
> udp6 0 0 ::1.123 *.*
> udp4 0 0 127.0.0.1.123 *.*
> udp6 0 0 fe80:1::2d0:59ff.123 *.*
> udp6 0 0 *.123 *.*
> udp4 0 0 *.123 *.*
>
> I don't get it. It's almost like it's trying to start twice, or forking
> at the wrong time, or something. Those ports for listening look
> pretty resonable, but it doesn't work, and it gives that error message.
>
> Very odd.
>
> It's definitely broke. Who wants to send in the PR?
Hi, I also had some problems with ntpd, when I upgraded to 5.3. I could solve
them by adding the "disable auth" option to my ntp.conf. At system startup I
run ntpdate before ntpd starts.
On the server ntp.matrix.net I run ntpd with the following config files (This
machine still runs 5.3-BETA-4):
# cat /etc/rc.conf | grep ntp
ntpdate_flags="-b clock.netcetera.dk tick.keso.fi"
ntpdate_enable="YES"
ntpd_enable="YES"
- -----------------
# cat /etc/ntp.conf
driftfile /var/db/ntpd.drift
# stratum 1
server swisstime.ethz.ch # 129.132.2.21
server time2.stupi.se # 192.36.143.151
server ntps1-0.cs.tu-berlin.de # 130.149.17.21
restrict 192.168.123.0 mask 255.255.255.0
disable auth
statsdir /var/log/ntpstats/
filegen peerstats file peerstats type day enable
filegen loopstats file loopstats type day enable
filegen clockstats file clockstats type day enable
- --------------------------------------------------
# ntpdc -c peers ntp.matrix.net
remote local st poll reach delay offset disp
=======================================================================
=swisstime.ee.et 84.101.100.100 1 1024 377 0.04387 -0.008958 0.01483
=Time2.Stupi.SE 84.101.100.100 1 1024 377 0.05922 -0.008854 0.01482
*hora.cs.tu-berl 84.101.100.100 1 1024 377 0.03352 -0.017613 0.01485
On the clients I run (these machines run 5.3-STABLE):
# cat /etc/rc.conf | grep ntp
ntpdate_flags="-b ntp.matrix.net"
ntpdate_enable="YES"
ntpd_enable="YES"
- -----------------
# cat /etc/ntp.conf
server ntp.matrix.net
restrict 127.0.0.1
statsdir /var/log/ntpstats/
filegen peerstats file peerstats type day enable
filegen loopstats file loopstats type day enable
filegen clockstats file clockstats type day enable
- --------------------------------------------------
# ntpdc -c peers
remote local st poll reach delay offset disp
=======================================================================
*ntp.matrix.net 192.168.123.10 2 1024 377 0.00046 -0.004711 0.01482
Cheers,
ch
- --
Christian Hiris <4711 at chello.at> | OpenPGP KeyID 0x3BCA53BE
OpenPGP-Key at hkp://wwwkeys.eu.pgp.net and http://pgp.mit.edu
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.6 (FreeBSD)
iD8DBQFB7HKV09WjGjvKU74RAkPwAJsFagaTCVXca0zTdYLiwJI1J63d5gCeKyO8
0NqA6Ve3YmlGls65zwBJxQE=
=mmJV
-----END PGP SIGNATURE-----
More information about the freebsd-questions
mailing list