Hi,
I've have an interesting time problem...
At work we have three classes of *nix computers:
1) Compaq DL360 G1 running Red Hat Linux 7
2) HP DL360 G4 running Red Hat Enterprise Linux 3
3) IBM System P AIX boxes running AIX 5.3L
All of them should have the same system time, but they
don't stay in sync. All machines are linked to the
same two ntp reference servers on-site. The reference time
servers on-site are Windows DC and they sync to another
Windows DC in California (USA) where our divisional head
office is.
The Red Hat Linux 7 systems are running ntpdate hourly (yes I
know that is bad).
The RHEL boxes are running a local NTPD.
The AIX boes are running IBM's xntpd.
Yesterday I forcibly set the clocks on all the Linux systems. They
then all appeared to be in sync with each other and the AIX boxes.
This morning the Linux systems running a manual ntpdate and the AIX
systems still synced, the RHEL systems running their own ntpd are now
drifting away from the other systems. Comparing these systems to my
Debian box at home which is running ntpd to pool.ntp.org, it looks
like the AIX boxes and manual Linux systems are okay, it's the RHEL
boxes running ntpd seem to be wrong.
One system running ntpd had gained 5 seconds since yesterday when
manually adjusted to the same ntp reference!
Any suggestions to fix the problem?
As ever thanks in advance?
--
Adam Trickett
Overton, HANTS, UK
A people that values its privileges above
its principles soon loses both.
-- Dwight D. Eisenhower
