Prefer Peer Plays Ping-Pong - NTP

This is a discussion on Prefer Peer Plays Ping-Pong - NTP ; Sorry, couldn't resist the subject line... I am using a HP Z3801A GPS-disciplined oscillator with both the hpgps and atom drivers in order to use the PPS signal as well as the serial data stream. I'm finding that the system ...

+ Reply to Thread
Results 1 to 5 of 5

Thread: Prefer Peer Plays Ping-Pong

  1. Prefer Peer Plays Ping-Pong

    Sorry, couldn't resist the subject line...

    I am using a HP Z3801A GPS-disciplined oscillator with both the hpgps
    and atom drivers in order to use the PPS signal as well as the serial
    data stream.

    I'm finding that the system keeps switching sync sources, like so:

    10 Sep 15:11:41 ntpd[42461]: synchronized to PPS(0), stratum 0
    10 Sep 15:13:51 ntpd[42461]: synchronized to GPS_HP(0), stratum 0
    10 Sep 15:14:54 ntpd[42461]: synchronized to PPS(0), stratum 0
    10 Sep 15:15:58 ntpd[42461]: synchronized to GPS_HP(0), stratum 0
    10 Sep 15:17:03 ntpd[42461]: synchronized to PPS(0), stratum 0
    10 Sep 15:19:10 ntpd[42461]: synchronized to GPS_HP(0), stratum 0

    I suspect this may be due to the (relatively) much greater jitter on the
    serial data stream. Here's the current billboard from ntpq -p:

    $ ntpq -p
    remote refid st t when poll reach delay offset jitter
    ================================================== ============================
    +GPS_HP(0) .GPS. 0 l 11 64 377 0.000 -0.620 3.171
    oPPS(0) .PPSZ. 0 l 9 64 255 0.000 0.001 0.004


    Here is my current config file (with logging and other irrelevant stuff
    pruned out):

    $ cat /etc/ntp.conf
    # /etc/ntp.conf for tock, using Z3801A and PPS

    enable pps

    # Z3801A
    # note: "ttl 1" sets comm params for Z3801A
    server 127.127.26.0 ttl 1 prefer
    fudge 127.127.26.0 time1 -0.960
    fudge 127.127.26.0 refid GPS

    # PPS
    server 127.127.22.0
    fudge 127.127.22.0 refid PPSZ
    fudge 127.127.22.0 flag3 1
    fudge 127.127.22.0 true

    Any suggestions on how I can force ntpd to lock to the PPS source and
    avoid this ping-pong effect? Any other suggestions on improving the
    configuration in general? (By the way -- this is on FreeBSD 5.4 and ntp
    4.2.0a).

    Thanks,

    John
    _______________________________________________
    questions mailing list
    questions@lists.ntp.isc.org
    https://lists.ntp.isc.org/mailman/listinfo/questions


  2. Re: Prefer Peer Plays Ping-Pong

    In article <450468C2.7050102@febo.com>,
    jra@febo.com (John Ackermann N8UR) writes:
    >Sorry, couldn't resist the subject line...
    >
    >I am using a HP Z3801A GPS-disciplined oscillator with both the hpgps
    >and atom drivers in order to use the PPS signal as well as the serial
    >data stream.
    >
    >I'm finding that the system keeps switching sync sources, like so:
    >
    >10 Sep 15:11:41 ntpd[42461]: synchronized to PPS(0), stratum 0
    >10 Sep 15:13:51 ntpd[42461]: synchronized to GPS_HP(0), stratum 0
    >10 Sep 15:14:54 ntpd[42461]: synchronized to PPS(0), stratum 0
    >10 Sep 15:15:58 ntpd[42461]: synchronized to GPS_HP(0), stratum 0
    >10 Sep 15:17:03 ntpd[42461]: synchronized to PPS(0), stratum 0
    >10 Sep 15:19:10 ntpd[42461]: synchronized to GPS_HP(0), stratum 0
    >
    >I suspect this may be due to the (relatively) much greater jitter on the
    >serial data stream. Here's the current billboard from ntpq -p:


    My system got a lot happier after I addded this line:
    tos mindist 0.010

    --
    The suespammers.org mail server is located in California. So are all my
    other mailboxes. Please do not send unsolicited bulk e-mail or unsolicited
    commercial e-mail to my suespammers.org address or any of my other addresses.
    These are my opinions, not necessarily my employer's. I hate spam.


  3. Re: Prefer Peer Plays Ping-Pong

    >>> In article , hmurray@suespammers.org (Hal Murray) writes:
    JA> I am using a HP Z3801A GPS-disciplined oscillator with both the hpgps and
    JA> atom drivers in order to use the PPS signal as well as the serial data
    JA> stream.

    JA> I'm finding that the system keeps switching sync sources, like so: ...
    JA> I suspect this may be due to the (relatively) much greater jitter on the
    JA> serial data stream. Here's the current billboard from ntpq -p:

    Hal> My system got a lot happier after I addded this line: tos mindist 0.010

    Please consider adding this to http://ntp.isc.org/Support, and I expect the
    site to be offline anytime between now (1:15pm PxT) and 4pm PxT due to
    maintenance on the building electrical system.

    H

  4. Re: Prefer Peer Plays Ping-Pong

    John,

    This problem has come up many times. The clock selection algorithm
    intersects the correctness interval to cleave the truechimers from the
    falsetickers. Problem is, the two correctness intervals sometimes do not
    overlap due to jitter. See the tos mindist command to set the minimum
    distance to something larger than 5 ms, say 10 ms.

    Dave

    John Ackermann N8UR wrote:
    > Sorry, couldn't resist the subject line...
    >
    > I am using a HP Z3801A GPS-disciplined oscillator with both the hpgps
    > and atom drivers in order to use the PPS signal as well as the serial
    > data stream.
    >
    > I'm finding that the system keeps switching sync sources, like so:
    >
    > 10 Sep 15:11:41 ntpd[42461]: synchronized to PPS(0), stratum 0
    > 10 Sep 15:13:51 ntpd[42461]: synchronized to GPS_HP(0), stratum 0
    > 10 Sep 15:14:54 ntpd[42461]: synchronized to PPS(0), stratum 0
    > 10 Sep 15:15:58 ntpd[42461]: synchronized to GPS_HP(0), stratum 0
    > 10 Sep 15:17:03 ntpd[42461]: synchronized to PPS(0), stratum 0
    > 10 Sep 15:19:10 ntpd[42461]: synchronized to GPS_HP(0), stratum 0
    >
    > I suspect this may be due to the (relatively) much greater jitter on the
    > serial data stream. Here's the current billboard from ntpq -p:
    >
    > $ ntpq -p
    > remote refid st t when poll reach delay offset jitter
    > ================================================== ============================
    > +GPS_HP(0) .GPS. 0 l 11 64 377 0.000 -0.620 3.171
    > oPPS(0) .PPSZ. 0 l 9 64 255 0.000 0.001 0.004
    >
    >
    > Here is my current config file (with logging and other irrelevant stuff
    > pruned out):
    >
    > $ cat /etc/ntp.conf
    > # /etc/ntp.conf for tock, using Z3801A and PPS
    >
    > enable pps
    >
    > # Z3801A
    > # note: "ttl 1" sets comm params for Z3801A
    > server 127.127.26.0 ttl 1 prefer
    > fudge 127.127.26.0 time1 -0.960
    > fudge 127.127.26.0 refid GPS
    >
    > # PPS
    > server 127.127.22.0
    > fudge 127.127.22.0 refid PPSZ
    > fudge 127.127.22.0 flag3 1
    > fudge 127.127.22.0 true
    >
    > Any suggestions on how I can force ntpd to lock to the PPS source and
    > avoid this ping-pong effect? Any other suggestions on improving the
    > configuration in general? (By the way -- this is on FreeBSD 5.4 and ntp
    > 4.2.0a).
    >
    > Thanks,
    >
    > John
    > _______________________________________________
    > questions mailing list
    > questions@lists.ntp.isc.org
    > https://lists.ntp.isc.org/mailman/listinfo/questions
    >



  5. Re: Prefer Peer Plays Ping-Pong

    >Hal> My system got a lot happier after I addded this line: tos mindist 0.010
    >
    >Please consider adding this to http://ntp.isc.org/Support, and I expect the
    >site to be offline anytime between now (1:15pm PxT) and 4pm PxT due to
    >maintenance on the building electrical system.


    There was a hard-to-find hint in the Z3801A refclock page.
    I expanded that to another section.

    Does this happen with other refclocks? If so, we should figure
    out where to add another page.


    --
    The suespammers.org mail server is located in California. So are all my
    other mailboxes. Please do not send unsolicited bulk e-mail or unsolicited
    commercial e-mail to my suespammers.org address or any of my other addresses.
    These are my opinions, not necessarily my employer's. I hate spam.


+ Reply to Thread