Public bug reported:
Binary package hint: ntpdate
two out of a few hundred times that I ran "ntpdate - u -s servernames"
my server seemed to have stopped his local clock.
Symptoms: answered ping and let root enter his name and password on
local console. But root was not logged in (no shell prompt ever
appeared), apache/dhcpd/sshd did not answer (at least not to an extend
that a user request was completed).
Judging the last log messages (that I generally show on one local console) the
clock time was not increased:
- Postfix logged "conenction limit reached". Postifx is usually bored on this
server. So I guess postfix opened connections but since the time stood still,
the maximum of connections per second/minute/whatever filled up.
- There were too many log messages in the same minute (I guess even the same
second, but can't remember, and log messages did not reach hard disk).
After reboot the BIOS setup showed the correct (universal coordinated)
time. So at least the mainboard clock seems not to be frozen.
In both cases the "clock freeze" happend in the minute when the cronjob
runs ntpdate (hourly):
/usr/sbin/ntpdate -u -s time1.rrzn.uni-hannover.de ntps1-1.cs.tu-
berlin.de timesrv1.uni-jena.de
package: ntpdate 1:4.2.4p4+dfsg-3ubuntu2.2 (distribution: Ubuntu
8.04.2)
kernel: 2.6.24-24.53-server (i686)
Mainboard: Dual Athlon MP board (Tyan Tiger mpx 2466N-4M) with AMD-760
MPX chipset (AMD-762 north bridge & AMD-768 south bridge), both CPUs
installed
** Affects: ntp (Ubuntu)
Importance: Undecided
Status: New
--
clock freezes when calling ntpdate (sometimes)
https://bugs.launchpad.net/bugs/379930
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
--
ubuntu-bugs mailing list
ubuntu-bugs@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
