[ntp:questions] Re: PPS kit and Spectracom TTS240rb
abrown at mlp.com
abrown at mlp.com
Tue Jul 11 20:24:02 UTC 2006
Ulrich Windl wrote:
> abrown at mlp.com writes:
> > Fellow techs,
> > Thanks for tuning into my issue of the day. I need to achieve 2 things.
> > The first is to obtain and install the PPS kit onto Spectracom's
> > TTS240rb. I spoke with the support folks a Spectracom and they
> > suggested I post here. The second is to install the PPS kernel
> > components on a RedHat ES 4 system running 2.6.9-XX.
> > I've managed to download and compile the nanokernel package from Mr.
> > David Mills' site, but getting the (working) hooks into my distro have
> > eluded me. The test program that was provided in the package ran
> > without incident (if that helps).
> So you are basically looking for someone that writes code for you? There is no
> released PPSkit for Linux kernel 2.6, and definitely not for any Enterprise
> Server release. Dave Mills' nanokernel is a "simulator" AFAIK/R. So what do
> you expect by running it?
> > Thanks in advance any and all help,
> > Allen
I was hoping the code was already written. My ultimate goal is to get
my clocks within 1ms.I've recently managed to get nearly all systems
consistenly within 5-20ms. In the introduction of "Generic Nanosecond
Kernel Timekeeping", they state that "...the primary purpose of these
modifications is to improve timekeeping accuracy to the order less than
a millisecond and, ultimately, to a nanosecond ...". In downloading the
code I wasn't entire sure what it included.
Regardless, after running the test scripts, it appeared that my kernel
and hardware could in fact get to within a millisecond. My next
question was, how do I get that improvement into my environment. The
answer from my vendor was 2 part: 1. get a PPS kit for my Spectracom
time server and 2. get the appropriate PPS implementation for my client
systems. In looking around the internet, I've found a few postings that
led me to believe that people have this working for other linux
implementations that were NTPv4 and various kernel revs.
Is this making sense? Thanks for your time.
More information about the questions