I'm not sure if my recollection is actually correct here but maybe this will
help some of you out a bit.

I'm digging back in my memory a bit but I remember dealing with a totally
runaway clock that would gain a several hours every few minutes.  (watching
the windows clock was fun)  It turned out that there's a calibration setting
in the PC that controls how many timer ticks are in one second (for
instance).  I think that there's a setting in some NTP clients that will
actually recalibrate your clock for you.  Caution: Something tells me that
this is what caused my runaway clock, it was inputting a larger and larger
correction each time.  So, I wouldn't leave the setting on all the time,
just turn it on, sync, calibrate and turn it off.  Then just use NTP
normally.

I just googled "clock drift" and there are some articles that describe the
problem.

Hope this helps,
Dave

On 3/6/06, Shidan <[EMAIL PROTECTED]> wrote:
>
> Nice, I can't even get my laptops clock to work properly!
>
> On 3/6/06, [EMAIL PROTECTED]
> <[EMAIL PROTECTED]> wrote:
> > Thanks, dude.  Excellent information on this list, as per usual.  Now if
> I could
> > just get the clock on the motherboard fixed. Seems like its losing its
> settings
> > a lot but the battery looks OK.  I think I have ntpd working OK in its
> place.
> >
> > Peter M.
> >
> > > There are definite benefits to a dual/multi CPU machine. The actual
> asterisk program isn't multi
> > > threaded so it won't utilize more then one but, when other processes
> kick off like transcoding,
> > > festival, comedian the OS will utilize the other CPU(s) to distribute
> the load. SO there is a definite
> > > benefit just not as much as one would totally want.
> > >
> > > The only time there is no benefit is when you have no transcoding and
> only the core asterisk
> > > process running. But this is highly unusual.
> > >
> > > Mike
> > >
> > > [EMAIL PROTECTED] wrote:
> > >     *This message was transferred with a trial version of
> CommuniGate(tm) Pro*
> > >     So running [EMAIL PROTECTED] on a dual processor P2 333 system is 
> > > still a
> waste of
> > >     processing power?  CentOS does recognize both processors and loads
> the
> > >     SMP kernel. Is there any benefit at all?
> > >
> > >     Peter M.
> > >
> > >
> > >     Maybe crazy enough that it will actually work. It amazes me
> sometimes
> > >     what ideas u come up with!! Some related news:
> > >
> > >     1) IAX is multithreaded in head now, so should work better on dual
> > >     processors than SIP, unless you're using the "other" asterisk sip
> > >     stack. Also,  a side benefit, silence suppression on IAX will
> probably
> > >     come soon.
> > >
> > >     On 3/2/06, Jim Van Meggelen <[EMAIL PROTECTED]> wrote:
> > >
> > >     Let me run something that's been floating about in my noggin by
> everyone:
> > >
> > >     Given that Asterisk does not make use of dual core CPUs or dual
> processors,
> > >     I was contemplating whether running Asterisk in two (or more)
> VMWare
> > >     sessions on a system might actually allow for more total
> performance. For
> > >     example, set up one VM to handle incoming lines, echo cancellation
> and all
> > >     sets, and then set up the other VM to handle VoIP, including
> transcoding.
> > >
> > >     A bit kludgy, to be sure, but would VMWare allow for both
> cores/CPUs to be
> > >     more fully utilized?
> > >
> > >     Very possibly not practical, but it's been floating about my head
> for a bit
> > >     and I figured I'd send it out into the ether to see what thoughts
> might come
> > >     back.
> > >
> > >     So . . . thoughts?
> > >
> > >     Jim.
> >
> > ********************************************************
> > Peter MacFarlane, ACP
> > Network Administration &  Programming
> > Target Call Center/ Message Centre P.E.I.
> > *****************************************************************
> > OpenBSD's PF Firewall: Now available with CARP Failover.
> > Nothing to do with fish, but everything to do with security!
> > *****************************************************************
> >
> >
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: [EMAIL PROTECTED]
> > For additional commands, e-mail: [EMAIL PROTECTED]
> >
> >
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
>
>


--
David Donovan
Consultant
Fulcrum Solutions

Reply via email to