You are not logged in.
Pages: 1
After rebooting from Windows, the time is off by 5:00 and nothing seems to fix it (short of a direct setting of the clock.) This is a very nagging issue, and I have gone through most of the posts regarding the issue with no luck. Here is the output of "timedatecrtl"
Local time: Sat, 2013-02-16 06:26:26 EST
Universal time: Sat, 2013-02-16 11:26:26 UTC
RTC time: Sat, 2013-02-16 11:26:26
Timezone: America/Detroit
UTC offset: -0500
NTP enabled: yes
NTP synchronized: yes
RTC in local TZ: no
DST active: no
Last DST change: EDT EST, DST became inactive
Sun, 2012-11-04 01:59:59 EDT
Sun, 2012-11-04 01:00:00 EST
Next DST change: EST EDT, DST will become active
the clock will jump one hour forward
Sun, 2013-03-10 01:59:59 EST
Sun, 2013-03-10 03:00:00 EDT
At the time of running the command the time was around 11:30 am EST. I am hoping NTPD sets the clock automatically after every boot as Windows seems to mess up the hardware clock. (I have fiddled with Windows side with no luck.) NTP does not seem to do anything even when doing "ntpd -qg" ...
Here is the status of ntpd
ntpd.service - Network Time Service
Loaded: loaded (/usr/lib/systemd/system/ntpd.service; enabled)
Active: active (running) since Sat, 2013-02-16 06:21:50 EST; 15min ago
Process: 1530 ExecStart=/usr/bin/ntpd -g -u ntp:ntp (code=exited, status=0/SUCCESS)
Main PID: 1531 (ntpd)
CGroup: name=systemd:/system/ntpd.service
`-1531 /usr/bin/ntpd -g -u ntp:ntp
Last edited by amiara (2013-02-16 16:38:37)
Offline
After rebooting from Windows, the time is off by 5:00 and nothing seems to fix it (short of a direct setting of the clock.) This is a very nagging issue, and I have gone through most of the posts regarding the issue with no luck. Here is the output of "timedatecrtl"
... RTC in local TZ: no ...
...
Set the hardware clock to local time zone
timedatectl set-local-rtc 1
Offline
You need to fix the issue, not band-aid it with ntp.
What version of Windows? Did you try the registry hack in the ArchWiki for telling Windows that the clock is in UTC?
Offline
@aesiris I have tried that and it does not have the desired effect. Besides I get that nice warning (This mode is not fully supported ...). It just sets RTC time to be the same as local time (the wrong local time ...), but "ntpd -qg" still has no effect.
@Scimmia I am using Windows 7 and have tried that registry hack "something like RTC is universal" ... As much as I like a cure, I am convinced that there is none and will be happy with a band-aid (aka a working ntp)
Last edited by amiara (2013-02-16 17:43:19)
Offline
How did you apply the registry change? If you entered it manually, are you positive you you didn't have any typos and created a dword entry? It's best to copy/paste the supplied reg file to avoid any of those issues. Did you also disable windows internet time sync?
NTP isn't working because the clock is too far off, it's a sanity check. Making ntp do the job is a bad thing, though, because it doesn't apply until the system is fully up. You end up with a situation where the system boots, then the clock goes way back in time, throwing everything off. Please don't do this.
If you absolutely must run Windows with a localtime RTC, you need to fix your linux setup to use it as well. Check /etc/adjtime, it should be three lines with the third one being "localtime".
Offline
@Scimmia: Thanks. I reapplied the registry change through the .reg file, and it has not yet taken effect. (I am not sure whether Windows decides to write down the clock at each shutdown or just some of the times ... we will see.) I have not disabled time sync in Windows as I don't see a reason. I want windows to be able to sync time.
My question is really why ntp has no effect. That -g switch in "ntpd -qg" is supposed to force it to change the clock, no matter how off it is (?). I don't want it to sanity check on my behalf. Is there a way to force ntp to set the clock? I suspect there is something wrong with my ntp setup ... I have set the time manually (say off by 1 minute) and expected it to get fine-tuned, but doesn't seem to happen.
Last edited by amiara (2013-02-16 20:12:45)
Offline
Sorry, I'm not going to help you abuse NTP. I already gave you everything you need to fix the real issues.
!bail
Offline
amiara, does /var/lib/ntp/ntp.drift exist and isn't empty ?
(it should have a number in it)
Disliking systemd intensely, but not satisfied with alternatives so focusing on taming systemd.
(A works at time B) && (time C > time B ) ≠ (A works at time C)
Offline
@Scimmia: Fair enough.
@Lone_Wolf: It has a number in it (7.496) Is there any diagnostics one can run on NTP to see what is going on?
Offline
If ntp starts succesfully, you should see something like this in the logs :
Feb 18 10:58:02 localhost ntpd[1207]: ntpd 4.2.6p5@1.2349-o Tue Dec 18 22:48:42 UTC 2012 (1)
Feb 18 10:58:02 localhost ntpd[1208]: proto: precision = 0.254 usec
Feb 18 10:58:02 localhost ntpd[1208]: ntp_io: estimated max descriptors: 1024, initial socket boundary: 16
Feb 18 10:58:02 localhost ntpd[1208]: Listen and drop on 0 v4wildcard 0.0.0.0 UDP 123
Feb 18 10:58:02 localhost ntpd[1208]: Listen and drop on 1 v6wildcard :: UDP 123
Feb 18 10:58:02 localhost ntpd[1208]: Listen normally on 2 lo 127.0.0.1 UDP 123
Feb 18 10:58:02 localhost ntpd[1208]: Listen normally on 3 eth0 192.168.178.2 UDP 123
Feb 18 10:58:02 localhost ntpd[1208]: Listen normally on 4 lo ::1 UDP 123
Feb 18 10:58:02 localhost ntpd[1208]: Listen normally on 5 eth0 fe80::230:48ff:febb:2454 UDP 123
Feb 18 10:58:02 localhost ntpd[1208]: Listen normally on 6 eth0 2001:980:d230:1:230:48ff:febb:2454 UDP 123
Feb 18 10:58:02 localhost ntpd[1208]: peers refreshed
Feb 18 10:58:02 localhost ntpd[1208]: Listening on routing socket on fd #23 for interface updates
ntpq -p gives a list of the peers ntp uses to keep the clock synchronised.
The jitter and offset vaues must not be zero, and 1 of the peers should have a * before it.
Disliking systemd intensely, but not satisfied with alternatives so focusing on taming systemd.
(A works at time B) && (time C > time B ) ≠ (A works at time C)
Offline
@Lone_Wolf, thanks. I forgot to check this thread for a while. The ntpq command is interesting. What if none of the peers has a star besides it? I have three of them and the jitter and offsets are nonzero for all of them.
Offline
The * is the server ntp will use first to synchronize, servers with + before them are second choice.
Servers with - before them are considered not stable enough (or your connection to them is not good enough).
I suggest you add some more servers in /etc/ntp.conf .
Here's my list for reference :
server ntp.xs4all.nl
server ntp2.xs4all.nl
server 0.nl.pool.ntp.org
server 1.nl.pool.ntp.org
server 2.nl.pool.ntp.org
server 3.nl.pool.ntp.org
The first 2 are ntp-servers run by my provider, the other 4 are chosen from the pool for the netherlands (the country where i live).
check http://support.ntp.org/bin/view/Servers/NTPPoolServers .
usually listing the country server pool and continent server pool works best. only add the worldwide pool if you can't get a good sync with country/continent pool.
Disliking systemd intensely, but not satisfied with alternatives so focusing on taming systemd.
(A works at time B) && (time C > time B ) ≠ (A works at time C)
Offline
Pages: 1