X-MimeOLE: Produced By Microsoft Exchange V6.5
Received: by onstor-exch02.onstor.net 
	id <01C86AD3.A778C1FA@onstor-exch02.onstor.net>; Fri, 8 Feb 2008 21:24:26 -0700
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Content-class: urn:content-classes:message
Subject: RE: How to reset NTP?
Date: Fri, 8 Feb 2008 21:24:25 -0700
Message-ID: <BB375AF679D4A34E9CA8DFA650E2B04E08351370@onstor-exch02.onstor.net>
In-Reply-To: <20080208194312.48bc6ea2@ripper.onstor.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: How to reset NTP?
Thread-Index: AchqzeowgzqIBfJMRgS8BwV1X8H2lgABXeiQ
References: <BB375AF679D4A34E9CA8DFA650E2B04E08350951@onstor-exch02.onstor.net><BB375AF679D4A34E9CA8DFA650E2B04E083509FC@onstor-exch02.onstor.net><BB375AF679D4A34E9CA8DFA650E2B04E08350A1E@onstor-exch02.onstor.net><BB375AF679D4A34E9CA8DFA650E2B04E08350D9B@onstor-exch02.onstor.net><BB375AF679D4A34E9CA8DFA650E2B04E083512F2@onstor-exch02.onstor.net><BB375AF679D4A34E9CA8DFA650E2B04E083512F6@onstor-exch02.onstor.net> <20080208194312.48bc6ea2@ripper.onstor.net>
From: "John Culp" <john.culp@onstor.com>
To: "Andy Sharp" <andy.sharp@onstor.com>,
	"Charissa Willard" <charissa.willard@onstor.com>
Cc: "dl-cstech" <dl-cstech@onstor.com>,
	"Joshua Goldenhar" <joshua.goldenhar@onstor.com>,
	"Rich LaReau" <rich.lareau@onstor.com>

ECR 22234 was filled for this.=20


John Culp=20
Regional SE Manager
940-239-7489 Office=20
972-523-4287 Cell=20
john.culp@onstor.com=20


-----Original Message-----
From: Andy Sharp=20
Sent: Friday, February 08, 2008 9:43 PM
To: Charissa Willard
Cc: John Culp; dl-cstech; Joshua Goldenhar; Rich LaReau
Subject: Re: How to reset NTP?

I believe that after a customer visit something like 9 months ago, I
filed a bug about setting ntp.  If I'm not mistaken, a change was made
where when you make a change to the ntp parameters, ntp is "refreshed"
or put another way, the system clock is jumped all at once to the ntp
reference, which also happens on bootup.

Does anyone else remember if that change ever went in?

Cheers,

a

On Fri, 8 Feb 2008 17:00:29 -0800 "Charissa Willard"
<charissa.willard@onstor.com> wrote:

> John,
>=20
> =20
>=20
> Please open a bug for this and mark it as an ECR. I'll see if we can=20
> get it in Kegg.
>=20
> =20
>=20
> Thanks,
>=20
> Charissa
>=20
> =20
>=20
> ________________________________
>=20
> From: John Culp
> Sent: Friday, February 08, 2008 4:56 PM
> To: Joshua Goldenhar; Rich LaReau; Charissa Willard
> Cc: dl-cstech
> Subject: RE: How to reset NTP?
>=20
> =20
>=20
> Is there a way that we could put a button in the GUI and a command in=20
> the code that would force a NTP refresh?  During install after setting

> the NTP the quickest way to force a refresh is to reboot the filer=20
> since modifying the NTP entry does not cause a synch.  This way if a=20
> time mismatch was discovered in a cluster we could force a refresh=20
> helping the Cluster DB get back in synch and installs a little easier.
>=20
> =20
>=20
> Just a thought.
>=20
> John Culp
> 940-239-7489 Office
> 972-523-4287 Cell
> john.culp@onstor.com
>=20
> =20
>=20
> =20
>=20
> ________________________________
>=20
> From: Joshua Goldenhar=20
> Sent: Friday, February 08, 2008 10:31 AM
> To: Rich LaReau; Charissa Willard
> Cc: dl-cstech
> Subject: RE: How to reset NTP?
>=20
> Is anyone filing a bug on this?
>=20
> =20
>=20
> It's nice that there is a fix, but the fact that this condition exists
> still seems like a bug. Or even the lack of a "ntp restart" command...
>=20
> =20
>=20
> -Josh=20
>=20
> Josh Goldenhar=20
> Phone: 408 963 2408, Cell: 408 547 7693=20
>=20
> =20
>=20
> 	=20
>=20
> =09
> ________________________________
>=20
>=20
> 	From: Rich LaReau=20
> 	Sent: Thursday, February 07, 2008 3:30 PM
> 	To: Charissa Willard
> 	Cc: dl-cstech
> 	Subject: RE: How to reset NTP?
>=20
> 	Thanks, Charissa!
>=20
> 	=20
>=20
> 	Yeah, the problem here is that everything looks good, and
> there are no errors seen in the logs.  Only that the time is... 8
> minutes off. We'll try to restart it and see if that fixes.
>=20
> 	=20
>=20
> 	Rich
>=20
> 	=20
>=20
> 	=20
>=20
> 	<nfxshCmd cmd=3D"system time show -v" rc=3D'0'>
> 	Thu Feb  7 13:29:34 CST 2008
> 	status=3D06f4 leap_none, sync_ntp, 15 events,
> event_peer/strat_chg,
> 	version=3D"ntpd 4.1.0 Tue Sep 18 19:27:43 PDT 2007 (1)",
> 	processor=3D"pmonmips", system=3D"OpenBSD2.8", leap=3D00,
> stratum=3D12, precision=3D-18, rootdelay=3D0.004, =
rootdispersion=3D38.455,
> peer=3D37573,
> 	refid=3D192.168.168.1,
>=20
> =09
> 	reftime=3Dcb55d8ba.ab02602c  Thu, Feb  7 2008 13:21:30.668,
> poll=3D10,
> 	clock=3Dcb55da9e.55a47a9e  Thu, Feb  7 2008 13:29:34.334,
> state=3D4,
>=20
> =09
> 	offset=3D2.000, frequency=3D-7.794, jitter=3D4.993, stability=3D0.013
> 	     remote           refid      st t when poll reach   delay
> offset  jitter
> =09
>
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
> =3D=3D=3D=3D=3D=3D
> 	 127.127.1.0     127.127.1.0     10 l   11   64  377    0.000
> 0.000   0.004
> 	*192.168.168.1   127.127.1.0     11 u  484 1024  377    0.004
> 2.000   3.343
> 	</nfxshCmd>
>=20
> 	=20
>=20
> 	=20
>=20
> 	=20
>=20
> =09
> ________________________________
>=20
>=20
> 	From: Charissa Willard=20
> 	Sent: Thursday, February 07, 2008 3:19 PM
> 	To: Rich LaReau
> 	Cc: dl-cstech
> 	Subject: RE: How to reset NTP?
>=20
> 	You can also restart the ntpd. If you list the processes on
> the filer, you will see a line like this:=20
>=20
> 	=20
>=20
> 	"/usr/local/sbin/ntpd -p /var/run/ntpd.pid".=20
>=20
> 	=20
>=20
> 	You can kill the current process and enter this line again to
> restart it. It's best to monitor the process and verify that the reach
> value eventually equals "377". You can monitor the ntpd process by
> entering "ntpq -q".=20
>=20
> 	=20
>=20
> 	Here's more information on ntp that explains the fields and
> all:
>=20
> 	=20
>=20
> 	1.) The "ntpdate -q <ip_address>" command returns info for the
> ntp server at the specified IP address:
>=20
> 	=20
>=20
> 	g8r11# ntpdate -q 10.0.0.18
>=20
> 	server 10.0.0.18, stratum 2, offset -0.220615, delay 0.04124
>=20
> 	 8 Jan 14:23:23 ntpdate[3190]: adjust time server 10.0.0.18
> offset -0.220615 sec
>=20
> 	=20
>=20
> 	2.) Use "ntptrace localhost" to trace back to the master time
> source:
>=20
> 	=20
>=20
> 	g8r11# ntptrace localhost
>=20
> 	localhost: stratum 3, offset 0.000000, synch distance 0.14552
>=20
> 	onstor-dc01.onstor.net: stratum 2, offset -0.210987, synch
> distance 0.08429
>=20
> 	ntp0.usno.navy.mil: stratum 1, offset -0.210468, synch
> distance 0.00046, refid 'USNO'
>=20
> 	=20
>=20
> 	3.) The "ntpq -p" command can be used to get the list of
> peers. (It's similar to part of what we display for "system time show
> -v".)=20
>=20
> 	=20
>=20
> 	g8r11# ntpq -p
>=20
> 	=20
>=20
> 	     remote           refid      st t when poll reach   delay
> offset  jitter
>=20
> =09
>
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
> =3D=3D=3D=3D=3D=3D
>=20
> 	 LOCAL(0)        LOCAL(0)        10 l   62   64  377    0.000
> 0.000   1.953
>=20
> 	*onstor-dc01.ons tock.usno.navy.  2 u  343  512  377    1.953
> -116.42  37.205
>=20
> 	=20
>=20
> 	*  address of the remote peer,=20
>=20
> 	*  reference ID (0.0.0.0 for an unknown reference ID),=20
>=20
> 	*  the stratum of the remote peer (a stratum of 16 indicates
> the remote peer is unsynchronized),=20
>=20
> 	*  type of the peer (local, unicast, multicast or broadcast),=20
>=20
> 	*  time the last packet was received, the polling interval
> (seconds),=20
>=20
> 	*  the polling interval (seconds),=20
>=20
> 	*  the reachability register (octal), (During a normal startup
> the registers values are these: 0, 1, 3, 7, 17, 37, 77, 177, 377.
> Example of an error: 257 (10101111) says that two valid responses were
> not received during the last eight polls. However, the last four polls
> worked fine.)
>=20
> 	*  the current estimated delay, offset and dispersion of the
> peer (seconds).
>=20
> 	=20
>=20
> 	The delay value is derived from the roundtrip time of the
> queries. The offset value shows the difference between the reference
> time and the system clock. The jitter value indicates the magnitude of
> jitter between several time queries. All values are milliseconds.
>=20
> 	The character in the left margin indicates the fate of this
> peer in the clock selection process:
>=20
> 	space
>=20
> 	discarded due to high stratum and/or failed sanity checks.
>=20
> 	x
>=20
> 	designated falseticker by the intersection algorithm.
>=20
> 	.
>=20
> 	culled from the end of the candidate list.
>=20
> 	-
>=20
> 	discarded by the clustering algorithm.
>=20
> 	+
>=20
> 	included in the final selection set.
>=20
> 	#
>=20
> 	selected for synchronization but distance exceeds maximum.
>=20
> 	*
>=20
> 	selected for synchronization.
>=20
> 	o
>=20
> 	selected for synchronization, pps signal in use.
>=20
> 	=20
>=20
> 	=20
>=20
> 	4.) Note that the ntp messages are also saved in the syslog
> message files:
>=20
> 	=20
>=20
> 	Jan  7 17:45:14 mktg3 ntpdate[14807]: step time server
> 10.0.0.18 offset 0.089707 sec
> 	Jan  7 17:45:14 mktg3 ntpd[20971]: ntpd 4.1.0 Fri Dec 28
> 21:31:16 PST 2007 (1)
> 	Jan  7 17:45:14 mktg3 ntpd[20971]: precision =3D 6 usec
> 	Jan  7 17:45:15 mktg3 ntpd[20971]: kernel time discipline
> status 0040
> 	Jan  7 17:45:15 mktg3 ntpd[20971]: frequency initialized
> -16.924 from /var/ntp/drift
>=20
> 	=20
>=20
> 	=20
>=20
> 	=20
>=20
> 	-----Original Message-----
> 	From: Rich LaReau=20
> 	Sent: Thursday, February 07, 2008 2:10 PM
> 	To: dl-cstech
> 	Subject: How to reset NTP?
>=20
> 	=20
>=20
> 	=20
>=20
> 	If it appears that NTP is not working on a node (or two)
> what's the recommended way to get it back?  I have a customer which
> shows that NTP appears to be configured and working, yet the two
> nodes are 2 and 5 minutes behind the NTP server, respectively. =20
>=20
> 	=20
>=20
> 	One recommendation is to delete and re-add the NTP server.  Is
> that the best option?
>=20
> 	=20
>=20
> 	Thanks,
>=20
> 	Rich
>=20
