21 Jan
2010
21 Jan
'10
2:41 p.m.
Hello,
I have just upgraded my VM from etch to lenny, and things look good overall.
One annoyance however is that on a reboot, the clock is ~30 mins in the
future and this is only fixed by a CRON @reboot job of "ntpdate", which
adjusts this correctly (offset -1935.350400 sec). However, this large
adjustment causes dovecot to kill itself.
On boot (via console) I can see :
..
drivers/rtc/hctosys.c: unable to open rtc device (rtc0)
..
Waiting for /dev to be fully populated...done.
Setting the system clock.
Unable to set System Clock to: Thu Jan 21 15:01:46 UTC 2010 (warning).
..
Setting the system clock.
Unable to set System Clock to: Thu Jan 21 15:01:49 UTC 2010 (warning).
Cleaning up ifupdown....
...
This is using lenny kernel 2.6.26-2-xen-686.
I tried setting "xen.independent_wallclock = 1" in /etc/sysctl.conf but this
does not make a difference.
hwclock --show
returns nothing.
Any idea why the clock needs such a large adjustment on reboot?
Cheers,
--
Alastair Sherringham
21 Jan
21 Jan
4 p.m.
Alastair Sherringham wrote:
Hello,
I have just upgraded my VM from etch to lenny, and things look good overall.
One annoyance however is that on a reboot, the clock is ~30 mins in the
future and this is only fixed by a CRON @reboot job of "ntpdate", which
adjusts this correctly (offset -1935.350400 sec). However, this large
adjustment causes dovecot to kill itself.
I'm not sure about Lenny/xen specific issues but the fact that machines
coming up can have deltas large enough to stop ntpd from adjusting the
clock is not uncommon, and normally you find people run ntpdate in
whatever rc file starts up ntpd so that ntp is able to adjust the time
from then on.
I'm surprised Lenny doesn't do this by default.
n
On boot (via console) I can see :
..
drivers/rtc/hctosys.c: unable to open rtc device (rtc0)
..
Waiting for /dev to be fully populated...done.
Setting the system clock.
Unable to set System Clock to: Thu Jan 21 15:01:46 UTC 2010 (warning).
..
Setting the system clock.
Unable to set System Clock to: Thu Jan 21 15:01:49 UTC 2010 (warning).
Cleaning up ifupdown....
...
This is using lenny kernel 2.6.26-2-xen-686.
I tried setting "xen.independent_wallclock = 1" in /etc/sysctl.conf but
this does not make a difference.
hwclock --show
returns nothing.
Any idea why the clock needs such a large adjustment on reboot?
Cheers,
--
Alastair Sherringham
------------------------------------------------------------------------
_______________________________________________
users mailing list
users(a)lists.bitfolk.com
https://lists.bitfolk.com/mailman/listinfo/users
4:14 p.m.
2010/1/21 Nigel Rantor <wiggly(a)wiggly.org>
I'm not sure about Lenny/xen specific issues but
the fact that machines
coming up can have deltas large enough to stop ntpd from adjusting the
clock is not uncommon, and normally you find people run ntpdate in
whatever rc file starts up ntpd so that ntp is able to adjust the time
from then on.
Thanks Nigel. Note - ntpdate is happy to adjust this - but dovecot thinks
this is bad and kills itself to prevent problems (
http://wiki.dovecot.org/TimeMovedBackwards).
This large time delta on boot is a big change from Debian Etch.
I am not running NTP - just ntpdate (cron - set clock once a day). Maybe I
shouldn't.
Cheers,
--
Alastair Sherringham
4:42 p.m.
Alastair Sherringham wrote:
2010/1/21 Nigel Rantor <wiggly(a)wiggly.org
<mailto:wiggly@wiggly.org>>
I'm not sure about Lenny/xen specific issues but the fact that machines
coming up can have deltas large enough to stop ntpd from adjusting the
clock is not uncommon, and normally you find people run ntpdate in
whatever rc file starts up ntpd so that ntp is able to adjust the time
from then on.
Thanks Nigel. Note - ntpdate is happy to adjust this - but dovecot
thinks this is bad and kills itself to prevent problems
(http://wiki.dovecot.org/TimeMovedBackwards).
Yeah, I understand, as long as the ntp rc script starts up before
dovecot then by the time dovecot is up and running things should be
fine. Lots of things get unhappy if the time apparently changes
underneath them.
This large time delta on boot is a big change from
Debian Etch.
Did you have ntpd running on that machine?
Either way, I'd still recommend ntpdate/ntp. For example, it also takes
acre of situations where, for example, a desktop machine's battery has
died and it is no longer keeping time when powered down.
I am not running NTP - just ntpdate (cron - set clock
once a day). Maybe
I shouldn't.
Ah, yes, that might do it. I would suggest you run ntpdate once only on
bootup to get the time correct so that ntp can then keep your clock from
slewing too far out of alignment from then on.
n
7:01 p.m.
Hi Alasdair,
On Thu, Jan 21, 2010 at 04:14:13PM +0000, Alastair Sherringham wrote:
Thanks Nigel. Note - ntpdate is happy to adjust this -
but dovecot
thinks this is bad and kills itself to prevent problems (
http://wiki.dovecot.org/TimeMovedBackwards).
This large time delta on boot is a big change from Debian Etch.
I'm sorry, I don't know why the host comes up with a large time
delta. I'll ask upstream and see if I can find out. On a normal
physical machine the host would take its system time from the
hardware clock, and most distributions set the hardware clock from
the system time when they shut down.
On a Xen domU using paravirtualization -- the current situation at
BitFolk -- there isn't any hardware clock. That's why there's a
message about failing to read the hardware clock during boot up.
That should be harmless though, because the host machine should
provide the domU with the correct time. The time is correct on all
the hosts; this is monitored.
As an aside, we can monitor your clock as well if you run an ntpd.
Just drop an email to support(a)bitfolk.com; monitoring of things with
Nagios is free.
I am not running NTP - just ntpdate (cron - set clock
once a day). Maybe I
shouldn't.
I would recommend that you do run ntpd because ntpdate will skew the
clock backwards if it needs to, and as you've seen many applications
don't like that.
ntpdate at (re)boot is a good idea.
Cheers,
Andy
--
http://bitfolk.com/ -- No-nonsense VPS hosting
7:22 p.m.
2010/1/21 Andy Smith <andy(a)bitfolk.com>
On Thu, Jan 21, 2010 at 04:14:13PM +0000, Alastair
Sherringham wrote:
Andy,
Looks like this has been a problem in the past. Digging briefly
(first/second hit) pinted me to a Debian bug + patch :
Bug#517808: dovecot-common: please call ntp-wait before starting Dovecot
http://www.mail-archive.com/debian-bugs-closed@lists.debian.org/msg218224.h…
Fixed in dovecot-common 1.1.11-3 (Squeeze/Backports). A patch to the init
script to do an ntp-wait.
I'll sort something out. Maybe use the backport repo.
Cheers,
--
Alastair Sherringham
Thanks Nigel. Note - ntpdate is happy to adjust
this - but dovecot
thinks this is bad and kills itself to prevent problems (
http://wiki.dovecot.org/TimeMovedBackwards).
This large time delta on boot is a big change from Debian Etch.
I'm sorry, I don't know why the host comes up with a large time
delta. I'll ask upstream and see if I can find out. On a normal
physical machine the host would take its system time from the
hardware clock, and most distributions set the hardware clock from
the system time when they shut down.
16 Feb
16 Feb
11:57 a.m.
On 21 January 2010 19:22, Alastair Sherringham <sherringham(a)gmail.com> wrote:
Looks like this has been a problem in the past.
Digging briefly
(first/second hit) pinted me to a Debian bug + patch :
Bug#517808: dovecot-common: please call ntp-wait before starting Dovecot
http://www.mail-archive.com/debian-bugs-closed@lists.debian.org/msg218224.h…
Fixed in dovecot-common 1.1.11-3 (Squeeze/Backports). A patch to the init
script to do an ntp-wait.
Sorry for following up so late. More information ...
To re-cap, I have had problems where the clock on my VP is 30 mins in
the future on boot. This behaviour started after a dist-upgrade to
Debian Lenny a few weeks ago. Correcting this (ntpdate,ntpd) causes
dovecot to die, and since Dovecot is what does AUTH for Postfix, both
Dovecot and Postfix are useless without a manual restart.
Linux calliope 2.6.26-2-xen-686 #1 SMP Wed Feb 10 11:28:19 UTC 2010
i686 GNU/Linux
This still happens and I have been testing all morning - rebooting the
VPS numerous times ...
ntpd wasn't helping
ntpdate wasn't helping
What now happens is that *sometimes* there is no clock skew. Sometimes
there is. Also - disabling ntpd (update-rc.d) and reinstalling ntpdate
has changed behaviour.
I now have ntpdate installed again - there is an
/etc/cron.daily/ntpdate containing :
ntpdate -s ntp0.sfo.bitfolk.com ntp0.lon.bitfolk.com
0.debian.pool.ntp.org 1.pool.ntp.org
There are no other crontab ntp calls (or @reboot jobs).
From boot, sometimes my clock is good, sometimes in the
future.
However, now, ntpdate *fixes* this before Dovecot dies, so I am
happier.
ntpdate[1036]: step time server 194.29.130.252 offset -2089.601967 sec
It is always ~30 mins.
I sit and reboot - login and check - reboot ...
OK no skew - dove OK
BAD - skew FIXED - dove OK
BAD - skew FIXED - dove OK
OK no skew - dove OK
OK no skew - dove OK (fsck ..)
BAD - skew FIXED - dove OK
OK no skew - dove OK
BAD - skew FIXED - dove OK
Odd?
Cheers,
--
Alastair Sherringham
6:28 p.m.
Hi Alastair,
On Tue, Feb 16, 2010 at 11:57:22AM +0000, Alastair Sherringham wrote:
To re-cap, I have had problems where the clock on my
VP is 30 mins in
the future on boot. This behaviour started after a dist-upgrade to
Debian Lenny a few weeks ago. Correcting this (ntpdate,ntpd) causes
dovecot to die, and since Dovecot is what does AUTH for Postfix, both
Dovecot and Postfix are useless without a manual restart.
Linux calliope 2.6.26-2-xen-686 #1 SMP Wed Feb 10 11:28:19 UTC 2010
i686 GNU/Linux
Are you using clocksource=jiffies on the kernel command line?
What is /proc/sys/xen/independent_wallclock set to?
Cheers,
Andy
--
http://bitfolk.com/ -- No-nonsense VPS hosting
6:57 p.m.
Andy Smith wrote:
Hi Alastair,
On Tue, Feb 16, 2010 at 11:57:22AM +0000, Alastair Sherringham wrote:
I think I have exactly the same problem, though I don't run dovecot,
hopefully Alastair doesn't mind me jumping in on his thread.
I don't seem to have clocksource=jiffies, and independent wallclock
seems to be 1:
joseph@button:~$ cat /boot/grub/menu.lst | grep -i jif
joseph@button:~$ cat /proc/sys/xen/independent_wallclock
1
I also infrequently get these messages on the console, no idea if
they're related or not:
[22433.110835] clocksource/0: Time went backwards: ret=6a3118eecc83d3
delta=-14067480 shadow=6a3118c913a34f offset=268faf29
Joseph
To re-cap, I have had problems where the clock on
my VP is 30 mins in
the future on boot. This behaviour started after a dist-upgrade to
Debian Lenny a few weeks ago. Correcting this (ntpdate,ntpd) causes
dovecot to die, and since Dovecot is what does AUTH for Postfix, both
Dovecot and Postfix are useless without a manual restart.
Linux calliope 2.6.26-2-xen-686 #1 SMP Wed Feb 10 11:28:19 UTC 2010
i686 GNU/Linux
Are you using clocksource=jiffies on the kernel command line?
What is /proc/sys/xen/independent_wallclock set to?
7:03 p.m.
Hello,
On Tue, Feb 16, 2010 at 06:57:08PM +0000, Joseph Heenan wrote:
I think I have exactly the same problem, though I
don't run dovecot,
hopefully Alastair doesn't mind me jumping in on his thread.
I don't seem to have clocksource=jiffies, and independent wallclock
seems to be 1:
The first thing I would try is having a kernel command line that
looks like this:
kernel /boot/vmlinuz-2.6.26-2-xen-686 root=UUID=d3249498-3e26-42c5-99a8-73fdd7ebcacb
clocksource=jiffies ro console=hvc0
along with /proc/sys/xen/independent_wallclock=1 and running ntpod,
this works fine for hundreds of Debian and Ubuntu VPSes running
2.6.2x Xen kernels.
Cheers,
Andy
--
http://bitfolk.com/ -- No-nonsense VPS hosting
<Ging> i stopped taking my medication long ago. the ironey is it was for
social anxioty disorder and now i daren't go back to the doctor to get
another perscription
7:18 p.m.
Andy Smith wrote:
Hello,
On Tue, Feb 16, 2010 at 06:57:08PM +0000, Joseph Heenan wrote:
Thanks, I've made that change.
I'm not entirely sure I am seeing the same thing as Alistair. At least
one of the things I see is two 'Unable to set System clock' messages
whilst the VPS is booting - though I now see they're happening before
independent_wallclock gets set:
Waiting for /dev to be fully populated...done.
Setting the system clock.
Unable to set System Clock to: Tue Feb 16 19:47:36 UTC 2010 (warning).
Activating swap:swapon on /dev/xvda2
[ 13.620255] Adding 262136k swap on /dev/xvda2. Priority:-1 extents:1
across:262136k
.
Will now check root file system:fsck 1.41.3 (12-Oct-2008)
[/sbin/fsck.ext3 (1) -- /] fsck.ext3 -a -C0 /dev/xvda1
/: clean, 43871/2293760 files, 1914874/4587520 blocks
.
[ 14.529561] EXT3 FS on xvda1, internal journal
Setting the system clock.
Unable to set System Clock to: Tue Feb 16 19:47:40 UTC 2010 (warning).
Cleaning up ifupdown....
Will now check all file systems.
fsck 1.41.3 (12-Oct-2008)
Checking all file systems.
Done checking file systems. A log is being saved in
/var/log/fsck/checkfs if that location is writable..
Setting kernel variables (/etc/sysctl.conf)...kernel.core_pattern =
/var/tmp/cores/core.%e.%p
xen.independent_wallclock = 1
done.
Will now mount local filesystems:.
The odd thing is the time was 19:16ish when this happened; I have no
idea where the 19:47 is coming from, but the ~30 minute difference does
seem consistent.
Joseph
I think I have exactly the same problem, though I
don't run dovecot,
hopefully Alastair doesn't mind me jumping in on his thread.
I don't seem to have clocksource=jiffies, and independent wallclock
seems to be 1:
The first thing I would try is having a kernel command line that
looks like this:
kernel /boot/vmlinuz-2.6.26-2-xen-686 root=UUID=d3249498-3e26-42c5-99a8-73fdd7ebcacb
clocksource=jiffies ro console=hvc0
along with /proc/sys/xen/independent_wallclock=1 and running ntpod,
this works fine for hundreds of Debian and Ubuntu VPSes running
2.6.2x Xen kernels.
7:34 p.m.
On 16 February 2010 19:18, Joseph Heenan <joseph(a)heenan.me.uk> wrote:
The odd thing is the time was 19:16ish when this
happened; I have no
idea where the 19:47 is coming from, but the ~30 minute difference does
seem consistent.
Look in /var/log/syslog - from a boot until the end. Mine starts 30
mins ahead until near the end, then jumps forward (assuming a "bad"
boot) :
e.g. see attached for an example - notice time jump.
I'll try Andy's jiffies, wallclock and ntpd suggestion ... reboot a
few times again :-)
--
Alastair Sherringham
7:41 p.m.
Alastair Sherringham wrote:
On 16 February 2010 19:18, Joseph Heenan
<joseph(a)heenan.me.uk> wrote:
ah, yes, I do see exactly the same kind of jump, even with 'jiffies' on
the command line:
Feb 16 19:47:45 button kernel: [ 20.057553] lo: Disabled Privacy
Extensions
Feb 16 19:47:49 button postfix/master[1239]: daemon started -- version
2.5.5, configuration /etc/postfix
Feb 16 19:12:58 button ntpdate[1030]: step time server 194.238.48.3
offset -2091.700794 sec
Joseph
The odd thing is the time was 19:16ish when this
happened; I have no
idea where the 19:47 is coming from, but the ~30 minute difference does
seem consistent.
Look in /var/log/syslog - from a boot until the end. Mine starts 30
mins ahead until near the end, then jumps forward (assuming a "bad"
boot) :
7:41 p.m.
Hi Joseph,
On Tue, Feb 16, 2010 at 07:18:37PM +0000, Joseph Heenan wrote:
I'm not entirely sure I am seeing the same thing
as Alistair. At least
one of the things I see is two 'Unable to set System clock' messages
whilst the VPS is booting - though I now see they're happening before
independent_wallclock gets set:
That's where the usual boot scripts attempt to read the hardware
clock (i.e. the BIOS) to get the system time from it. Xen
paravirtual machines don't have hardware clocks so this always
fails, and is harmless.
As long as you have xen.independent_wallclock set to 1 then you
should be able to set your clock yourself. If you don't then you
won't be able to change it at all. It will just fail silently.
I recommend that you set that to 1, run ntpd and put in a support
ticket to support(a)bitfolk.com for me to monitor your ntpd for you.
You'll know then if it ever loses sync.
I do find that the clock is sometimes set oddly after boot or after
a suspend/restore, but I am able to fix it with ntpdate and keep it
synced with ntpd.
Cheers,
Andy
--
http://bitfolk.com/ -- No-nonsense VPS hosting
8:11 p.m.
On 16 February 2010 19:03, Andy Smith <andy(a)bitfolk.com> wrote:
The first thing I would try is having a kernel command
line that
looks like this:
kernel /boot/vmlinuz-2.6.26-2-xen-686 root=UUID=d3249498-3e26-42c5-99a8-73fdd7ebcacb
clocksource=jiffies ro console=hvc0
along with /proc/sys/xen/independent_wallclock=1 and running ntpod,
this works fine for hundreds of Debian and Ubuntu VPSes running
2.6.2x Xen kernels.
Andy,
It is still going wrong - 4th reboot - and this time, Dovecot died as
well - at least previously, Dovecot seemed to stay running.
OK
OK
OK
BAD
I had to do a "mknod" for xvda to update grub.
This is what I did :
1) /boot/grub/menu.lst :
Add "clocksource=jiffies" to kopt - run update-grub
Now :
/sys/devices/system/clocksource/clocksource0/current_clocksource
is "jiffies"
2) /etc/sysctl.conf
Add :
xen.independent_wallclock=1
/proc/sys/xen/independent_wallclock is 1
3) Re-enable ntpd
update-rd.d ntp defaults
I confirm all the above is set.
Very strange.
--
Alastair Sherringham
8:47 p.m.
On 16 February 2010 20:16, Alastair Sherringham <sherringham(a)gmail.com> wrote:
Here's the syslog from the bad reboot.
I purged ntpdate, just in case - but still a problem. Attached syslog again.
ntpd[1247]: time reset -2092.079691 s
(but Dovecot starts at least)
--
Alastair Sherringham
8:57 p.m.
Hi Alastair,
On Tue, Feb 16, 2010 at 08:16:36PM +0000, Alastair Sherringham wrote:
Of
course not. Everyone knows you can't have your kayak and heat it.
-- James Fidell
Feb 16 20:35:13 calliope kernel: [ 47.073409] eth0:
no IPv6 routers present
Feb 16 20:00:28 calliope ntpdate[1043]: step time server 192.33.96.102 offset
-2091.895883 sec
I appreciate that time went backwards here which upsets dovecot,
but..
Feb 16 20:00:37 calliope ntpd[1187]: synchronized to
213.132.202.192, stratum 2
You're now synced to an NTP server and have the correct time. Does
it later lose sync?
Is it possible that what you describe as an "OK" boot had
independent_wallclock set to 0 so you were unable to set the time,
it remained wrong, and dovecot did not complain? Personally I would
rather have the right time and make dovecot start after ntpdate.
Having it boot with the wrong time and then have ntpdate/ntp adjust
it is the best I can offer I'm afraid, because that's the best I've
been able to achieve.
Just in case you might be thinking that the host you're on has the
wrong time:
kahlua$ date && ntpq -np
Tue Feb 16 20:47:13 UTC 2010
remote refid st t when poll reach delay offset jitter
==============================================================================
+209.237.247.192 212.13.195.4 4 u 770 1024 377 140.306 0.702 0.091
212.13.195.3 212.13.204.4 3 u 900 1024 377 0.280 2.328 0.271
212.13.195.4 212.13.204.4 3 u 910 1024 377 0.364 1.133 7.864
2001:ba8:1f1:f0 212.13.204.4 3 u 764 1024 377 4.560 0.207 1.500
2001:ba8:1f1:f0 212.13.204.4 3 u 767 1024 377 0.336 0.264 0.739
*212.13.204.4 195.66.241.10 2 u 29 1024 377 0.496 2.619 1.066
+212.13.204.5 195.66.241.2 2 u 989 1024 377 0.559 -0.330 0.066
So I don't know why it would be consistently around 35 minutes off.
I do note that:
kahlua$ date && hwclock --show && uptime
Tue Feb 16 20:49:59 UTC 2010
Tue 16 Feb 2010 21:08:23 UTC -0.710159 seconds
20:50:00 up 349 days, 2:02, 5 users, load average: 0.52, 0.89, 1.23
which is obviously about 20 minutes off, but the hardware clock will
get set from the system clock next time it's shut down, and I can't
imagine why it would be used for anything. I'll sync it from the
system clock now, though.
kahlua$ sudo hwclock --systohc && date && hwclock --show
Password:
Tue Feb 16 20:51:53 UTC 2010
Tue 16 Feb 2010 20:51:54 UTC -0.472518 seconds
Does that change things?
Cheers,
Andy
--
http://bitfolk.com/ -- No-nonsense VPS hosting
I have just
recently purchased a Feathercraft Big Kahuna kayak
does it have a heater?
9:46 p.m.
On 16 February 2010 20:57, Andy Smith <andy(a)bitfolk.com> wrote:
Hello Andy,
No, as far as I can see, it's only the initial boot that starts 30
mins out. Once we sync, we're OK.
Feb 16
20:00:37 calliope ntpd[1187]: synchronized to 213.132.202.192, stratum 2
You're now synced to an NTP server and have the correct time. Does
it later lose sync? Is it possible that what you describe as an
"OK" boot had
independent_wallclock set to 0 so you were unable to set the time,
it remained wrong, and dovecot did not complain? Personally I would
rather have the right time and make dovecot start after ntpdate.
No - the "bad" boot was always the time jumping 30 mins - the time has
always ended up correct after boot, just sometimes needing the large
adjustment. which sometimes makes Dovecot die.
Having it boot with the wrong time and then have
ntpdate/ntp adjust
it is the best I can offer I'm afraid, because that's the best I've
been able to achieve.
That's fine.
I am OK if I can find a setting that ends up with the right time
(whether we start right or need a large adjustment) and doesn't cause
Dovecot to die. I think I can get to this state (if I am not already).
Just odd - sometimes 30 mins out at start, sometimes not.
kahlua$ sudo hwclock --systohc && date
&& hwclock --show
Password:
Tue Feb 16 20:51:53 UTC 2010
Tue 16 Feb 2010 20:51:54 UTC -0.472518 seconds
Does that change things?
I'll do some more tests tomorrow and report back anything of interest.
Thanks for the help.
Cheers,
--
Alastair Sherringham
17 Feb
17 Feb
9:27 a.m.
Andy Smith wrote:
Having it boot with the wrong time and then have
ntpdate/ntp adjust
it is the best I can offer I'm afraid, because that's the best I've
been able to achieve.
Ah, okay - so long as I've not broken anything then I'm happy :-)
(and I've not seen the weird kern.log message since adding the
clocksource=jiffies.)
Thanks
Joseph
16 Feb
16 Feb
7:26 p.m.
On 16 February 2010 18:28, Andy Smith <andy(a)bitfolk.com> wrote:
Are you using clocksource=jiffies on the kernel
command line?
No.
# cat /sys/devices/system/clocksource/clocksource0/current_clocksource
xen
The other option is "jiffies", which I'd not normally want I think.
What is /proc/sys/xen/independent_wallclock set to?
That's not set. But I did add to sysctl.conf a while back to test -
commented out today though - I don't think it made a difference.
Cheers,
--
Alastair Sherringham
21 Jan
21 Jan
4:22 p.m.
2010/1/21 Alastair Sherringham <sherringham(a)gmail.com>om>:
Any idea why the clock needs such a large adjustment
on reboot?
I have no idea whatsoever, but I wonder what your kernel boot
parameters are? My (lenny) VM has clocksource=jiffies - maybe this
has something to do with it?
A quick Google search says that at least one person somewhere reported
Good Things happening after this.
Maybe this wasn't the default for Etch VMs and wasn't migrated?
Cheers,
Lee
5:37 p.m.
2010/1/21 Alastair Sherringham <sherringham(a)gmail.com>
Any idea why the clock needs such a large adjustment
on reboot?
Thanks for the replies. My main concern was with the large adjustment and
whether I had a misconfiguration. I'll work around the dovecot issue as per
their wiki.
Cheers,
--
Alastair Sherringham
5421
days inactive
5448
days old
22 comments
5 participants
participants (5)
-
Alastair Sherringham -
Andy Smith -
Joseph Heenan -
Lee Gent -
Nigel Rantor