RT causing time to drift?

RT Developers
1

Has anyone noticed any problems with time drift? I have RT and RTIR
loaded onto Fedora Core 4 and the time drifts horribly. I have set
the date manually and have had it drift immediately. I also notice
that something is trying to call set_rmc_mmss in order to set the
system time to the hardware clock. I’ve tried using ntp time but for
some reason I can’t get any of the servers to connect. The servers
are all specified and show up when i do an ntpq -p. However, there is
no asterisk next to any of the servers indicating a relationship.

I’ve loaded Fedora Core 4 before, enabled my firewall rules, and
apache, mysql, and a slew of other things and the time using ntp has
been rock solid. Could one of the perl modules be acting up?

Thanks,
Alex

smime.p7s (2.32 KB)

2

Has anyone noticed any problems with time drift? I have RT and RTIR
loaded onto Fedora Core 4 and the time drifts horribly. I have set
the date manually and have had it drift immediately. I also notice
that something is trying to call set_rmc_mmss in order to set the
system time to the hardware clock. I’ve tried using ntp time but for
some reason I can’t get any of the servers to connect. The servers
are all specified and show up when i do an ntpq -p. However, there is
no asterisk next to any of the servers indicating a relationship.

I’ve loaded Fedora Core 4 before, enabled my firewall rules, and
apache, mysql, and a slew of other things and the time using ntp has
been rock solid. Could one of the perl modules be acting up?

The only way RT is involved is if it is causing heavy load
and heavy load is the cause.

-Todd

3

To answer a few questions that have cropped up:

I am running it on Fedora Core 4 loaded onto a VMware instance.
Perhaps this is a nono in the RT/RTIR world, I didn’t read anything
saying otherwise. The NTP holes are punched through on the firewall,
I’ve even gone to such drastic measures as flushing all the iptables
rules and completely shutting iptables down (non-production machine
obviously).

Thanks,
AlexOn Dec 14, 2005, at 5:45 PM, Alex Meyer wrote:

Has anyone noticed any problems with time drift? I have RT and RTIR
loaded onto Fedora Core 4 and the time drifts horribly. I have set
the date manually and have had it drift immediately. I also notice
that something is trying to call set_rmc_mmss in order to set the
system time to the hardware clock. I’ve tried using ntp time but
for some reason I can’t get any of the servers to connect. The
servers are all specified and show up when i do an ntpq -p.
However, there is no asterisk next to any of the servers indicating
a relationship.

I’ve loaded Fedora Core 4 before, enabled my firewall rules, and
apache, mysql, and a slew of other things and the time using ntp
has been rock solid. Could one of the perl modules be acting up?

Thanks,
Alex

Rt-devel mailing list
Rt-devel@lists.bestpractical.com
The rt-devel Archives

smime.p7s (2.32 KB)

4

I’ve loaded Fedora Core 4 before, enabled my firewall rules, and
apache, mysql, and a slew of other things and the time using ntp
has been rock solid. Could one of the perl modules be acting up?

That’s funny… the clock is maintained by the OS. Unless you’re
running your web server as root there’s no way it can affect it.

There is such a thing as bad hardware, too… I’d run your vendor
hardware diagnostics and see what you can find.

smime.p7s (2.42 KB)

5

I am running it on Fedora Core 4 loaded onto a VMware instance.

VMware is known to play fast and loose with clock interrupts. Never
have I seen an accurate clock under VMware except when running windows.

smime.p7s (2.42 KB)