You are not logged in.
Hi,
I upgraded to kernel 2.6.37 64 bit this night with no error messages.
But i run into a serious issue, i run an NTP server and after the kernel upgrade my time is totally screwed up.
NTPD is not syncing and jitter i > 6 for all servers.
I disabled IPv6 just in case and that does not help.
The offsett seems to increase.
Does anybody have the same problem?
ntpq -p
remote refid st t when poll reach delay offset jitter
==============================================================================
+ntp-public.uit. 129.242.6.241 2 u 40 64 37 16.905 -15.175 8.583
+ns1.uio.no 158.38.0.237 2 u 35 64 37 10.094 -11.508 6.378
-2.81-166-42.cus 193.79.237.14 2 u 39 64 37 16.224 -0.737 11.265
+fartein.ifi.uio 195.220.94.163 2 u 40 64 37 10.807 -13.231 8.292
+ntp4.interpost. 139.112.7.52 2 u 34 64 37 12.667 -10.371 6.201
*213.75.60.245 .PPS. 1 u 33 64 37 32.389 -16.970 9.642
+ntp1.nl.uu.net .PPS. 1 u 31 64 37 38.092 -6.723 6.194
ntpdate
ntpdate[1568]: step time server 81.166.42.2 offset -3599.879091 sec
Last edited by lars_4 (2011-03-02 23:16:04)
Offline
I have tried clocksource=acpi_pm, clocksource=pit and tsc and it does not make a difference.
Offline
lars_4, please remove the URGENT tag from your topic title. See: How to Post.
ᶘ ᵒᴥᵒᶅ
Offline
lars_4, please remove the URGENT tag from your topic title. See: How to Post.
I removed it..
Last edited by lars_4 (2011-03-02 23:33:05)
Offline
Being rude to the mods is a very effective way to get your thread closed.
You should try removing '/var/lib/hwclock/adjtime' to reset your time skew.
Steven [ web : git ]
GPG: 327B 20CE 21EA 68CF A7748675 7C92 3221 5899 410C
Do not email: honeypot@stebalien.com
Offline
Being rude to the mods is a very effective way to get your thread closed.
You should try removing '/var/lib/hwclock/adjtime' to reset your time skew.
I checked my timezones: date command = hwclock -r
/etc/rc.conf timezone is correct, HARDWARECLOCK = "localtime"
remove the adjtime file and rebooted, no effect
Offline
I'd say the jitter would be caused by wildly varying round trip times, as in having heavy network usage or the time servers being on a network with heavy network usage.
However (and I'm in no way an expert) I'd say that the jitter should only affect the precision you have when syncing the clock and it should not affect drift.
Another thing is that if you are running only linux you might want to set HARDWARECLOCK="UTC" but ask for more input about that.
R00KIE
Tm90aGluZyB0byBzZWUgaGVyZSwgbW92ZSBhbG9uZy4K
Offline
remote refid st t when poll reach delay offset jitter
==============================================================================
*ns1.uio.no 158.38.0.237 2 u 41 64 377 10.411 52.520 21.274
+2.81-166-42.cus 193.79.237.14 2 u 36 64 377 16.268 34.235 14.173
+ntp1.interpost. 139.117.105.71 2 u 37 64 377 12.222 47.244 17.513
Reset the hardwareclock, removed adjtime file, re-installed ntpd, rebooted cisco router, tried localtime and UTC in /etc/rc.conf.
I am also using NOTSC as boot parameter, seems to be a little more stable. (using HPET now)
Any more suggentions?
Last edited by lars_4 (2011-03-03 17:35:58)
Offline
If you're unit is a laptop...
Did you installed by any change those "laptop-mode-tools"?
I've installed those once and my hw-clock starts running ahead of time (10 min plus a day)...
Offline