linuxptp-users

[Linuxptp-users] system1: ptp4l master offset continues to climb: step_threshold 1

[Harold L. <hla...@pi...>]

To Whom It May Concern,

Have a multi-system configuration and on one of the systems 'ptp4l master offset' continues to climb with 'step_theshold 1', might there be some other configuration parameter I am over-looking ?

Thanks,
Harold

system:

               ptp4l.conf:

                              root@zx3-pm3-zynq7:~# cat /usr/bin/ptp4l.conf
                              [global]
                              verbose 1
                              step_threshold 1
                              use_syslog 0
                              [eth0]


               ptp4l:

                              root@zx3-pm3-zynq7:~# /usr/bin/ptp4l -f /usr/bin/ptp4l.conf &

                              ptp4l[4477.362]: port 1: INITIALIZING to LISTENING on INITIALIZE
                              ptp4l[4477.363]: port 0: INITIALIZING to LISTENING on INITIALIZE
                              ptp4l[4478.668]: port 1: new foreign master 000a35.fffe.01225c-1
                              ptp4l[4482.668]: selected best master clock 000a35.fffe.01225c
                              ptp4l[4482.668]: port 1: LISTENING to UNCALIBRATED on RS_SLAVE
                              ptp4l[4484.392]: master offset   -2756145 s0 freq -3910000 path delay     30485
                              ptp4l[4485.393]: master offset   -2767595 s1 freq -3910000 path delay     30485
                              ptp4l[4486.393]: master offset     -11892 s2 freq -3910000 path delay     30485
                              ptp4l[4486.393]: port 1: UNCALIBRATED to SLAVE on MASTER_CLOCK_SELECTED
                              ptp4l[4487.393]: master offset     -23642 s2 freq -3910000 path delay     30485
                              ptp4l[4488.393]: master offset     -34899 s2 freq -3910000 path delay     30485
                              ptp4l[4489.393]: master offset     -47149 s2 freq -3910000 path delay     30485
                              ptp4l[4490.393]: master offset     -57433 s2 freq -3910000 path delay     29612
                              ptp4l[4491.394]: master offset     -69590 s2 freq -3910000 path delay     29368
                              ptp4l[4492.394]: master offset     -81961 s2 freq -3910000 path delay     29612
                              ptp4l[4493.394]: master offset     -93790 s2 freq -3910000 path delay     29612
                              ptp4l[4494.394]: master offset    -106627 s2 freq -3910000 path delay     29368



               phc2sys:

                              phc2sys[4598.771]: port 00214a.fffe.000003-1 changed state
                              phc2sys[4598.772]: reconfiguring after port state change
                              phc2sys[4598.773]: selecting CLOCK_REALTIME for synchronization
                              phc2sys[4598.773]: selecting eth0 as the master clock
                              phc2sys[4598.773]: phc offset -10536165 s0 freq   +3833 delay   2814
                              phc2sys[4599.774]: phc offset -10517174 s1 freq  +22810 delay   2838
                              phc2sys[4600.774]: phc offset    -22424 s2 freq    +386 delay   2769
                              phc2sys[4601.775]: phc offset    -22820 s2 freq   -6737 delay   2841
                              phc2sys[4602.775]: phc offset    -16038 s2 freq   -6801 delay   2787
                              phc2sys[4603.775]: phc offset     -9142 s2 freq   -4717 delay   2829
                              phc2sys[4604.776]: phc offset     -4444 s2 freq   -2761 delay   2802
                              phc2sys[4605.776]: phc offset     -1673 s2 freq   -1324 delay   2817
                              phc2sys[4606.777]: phc offset      -312 s2 freq    -465 delay   2775
                              phc2sys[4607.777]: phc offset       149 s2 freq     -97 delay   2835
                              phc2sys[4608.778]: phc offset       297 s2 freq     +96 delay   2802
                              phc2sys[4609.778]: phc offset       197 s2 freq     +85 delay   2799
                              phc2sys[4610.778]: phc offset       137 s2 freq     +84 delay   2790
                              phc2sys[4611.779]: phc offset        63 s2 freq     +51 delay   2811
                              phc2sys[4612.779]: phc offset        49 s2 freq     +56 delay   2754
                              phc2sys[4613.780]: phc offset       -20 s2 freq      +1 delay   2814
                              phc2sys[4614.780]: phc offset        35 s2 freq     +50 delay   2787
                              phc2sys[4615.441]: phc offset        -7 s2 freq     +19 delay   2754

Re: [Linuxptp-users] system1: ptp4l master offset continues to climb: step_threshold 1

[Richard C. <ric...@gm...>]

On Thu, Sep 17, 2015 at 03:13:55PM +0000, Harold Lapprich wrote:
> ptp4l[4486.393]: port 1: UNCALIBRATED to SLAVE on MASTER_CLOCK_SELECTED
> ptp4l[4487.393]: master offset     -23642 s2 freq -3910000 path delay     30485
> ptp4l[4488.393]: master offset     -34899 s2 freq -3910000 path delay     30485
> ptp4l[4489.393]: master offset     -47149 s2 freq -3910000 path delay     30485
> ptp4l[4490.393]: master offset     -57433 s2 freq -3910000 path delay     29612
> ptp4l[4491.394]: master offset     -69590 s2 freq -3910000 path delay     29368
> ptp4l[4492.394]: master offset     -81961 s2 freq -3910000 path delay     29612
> ptp4l[4493.394]: master offset     -93790 s2 freq -3910000 path delay     29612
> ptp4l[4494.394]: master offset    -106627 s2 freq -3910000 path delay     29368

Looks like the xilinx driver and/or hw is broken.

Sorry,
Richard

Re: [Linuxptp-users] system1: ptp4l master offset continues to climb: step_threshold 1

[Dale S. <dal...@gm...>]

I could be wrong, but isn't the step_threshold in seconds?  And -106627 is
about  106 microseconds.  So won't it take quite a while yet before the
threshold is reached?

-Dale

On Thu, Sep 17, 2015 at 11:13 AM, Harold Lapprich <
hla...@pi...> wrote:

> To Whom It May Concern,
>
>
>
> Have a multi-system configuration and on one of the systems 'ptp4l master
> offset' continues to climb with 'step_theshold 1', might there be some
> other configuration parameter I am over-looking ?
>
>
>
> Thanks,
>
> Harold
>
>
>
> system:
>
>
>
>                ptp4l.conf:
>
>
>
>                               root@zx3-pm3-zynq7:~# cat
> /usr/bin/ptp4l.conf
>
>                               [global]
>
>                               verbose 1
>
>                               step_threshold 1
>
>                               use_syslog 0
>
>                               [eth0]
>
>
>
>
>
>                ptp4l:
>
>
>
>                               root@zx3-pm3-zynq7:~# /usr/bin/ptp4l -f
> /usr/bin/ptp4l.conf &
>
>
>
>                               ptp4l[4477.362]: port 1: INITIALIZING to
> LISTENING on INITIALIZE
>
>                               ptp4l[4477.363]: port 0: INITIALIZING to
> LISTENING on INITIALIZE
>
>                               ptp4l[4478.668]: port 1: new foreign master
> 000a35.fffe.01225c-1
>
>                               ptp4l[4482.668]: selected best master clock
> 000a35.fffe.01225c
>
>                               ptp4l[4482.668]: port 1: LISTENING to
> UNCALIBRATED on RS_SLAVE
>
>                               ptp4l[4484.392]: master offset   -2756145 s0
> freq -3910000 path delay     30485
>
>                               ptp4l[4485.393]: master offset   -2767595 s1
> freq -3910000 path delay     30485
>
>                               ptp4l[4486.393]: master offset     -11892 s2
> freq -3910000 path delay     30485
>
>                               ptp4l[4486.393]: port 1: UNCALIBRATED to
> SLAVE on MASTER_CLOCK_SELECTED
>
>                               ptp4l[4487.393]: master offset     -23642 s2
> freq -3910000 path delay     30485
>
>                               ptp4l[4488.393]: master offset     -34899 s2
> freq -3910000 path delay     30485
>
>                               ptp4l[4489.393]: master offset     -47149 s2
> freq -3910000 path delay     30485
>
>                               ptp4l[4490.393]: master offset     -57433 s2
> freq -3910000 path delay     29612
>
>                               ptp4l[4491.394]: master offset     -69590 s2
> freq -3910000 path delay     29368
>
>                               ptp4l[4492.394]: master offset     -81961 s2
> freq -3910000 path delay     29612
>
>                               ptp4l[4493.394]: master offset     -93790 s2
> freq -3910000 path delay     29612
>
>                               ptp4l[4494.394]: master offset    -106627 s2
> freq -3910000 path delay     29368
>
>
>
>
>
>
>
>                phc2sys:
>
>
>
>                               phc2sys[4598.771]: port 00214a.fffe.000003-1
> changed state
>
>                               phc2sys[4598.772]: reconfiguring after port
> state change
>
>                               phc2sys[4598.773]: selecting CLOCK_REALTIME
> for synchronization
>
>                               phc2sys[4598.773]: selecting eth0 as the
> master clock
>
>                               phc2sys[4598.773]: phc offset -10536165 s0
> freq   +3833 delay   2814
>
>                               phc2sys[4599.774]: phc offset -10517174 s1
> freq  +22810 delay   2838
>
>                               phc2sys[4600.774]: phc offset    -22424 s2
> freq    +386 delay   2769
>
>                               phc2sys[4601.775]: phc offset    -22820 s2
> freq   -6737 delay   2841
>
>                               phc2sys[4602.775]: phc offset    -16038 s2
> freq   -6801 delay   2787
>
>                               phc2sys[4603.775]: phc offset     -9142 s2
> freq   -4717 delay   2829
>
>                               phc2sys[4604.776]: phc offset     -4444 s2
> freq   -2761 delay   2802
>
>                               phc2sys[4605.776]: phc offset     -1673 s2
> freq   -1324 delay   2817
>
>                               phc2sys[4606.777]: phc offset      -312 s2
> freq    -465 delay   2775
>
>                               phc2sys[4607.777]: phc offset       149 s2
> freq     -97 delay   2835
>
>                               phc2sys[4608.778]: phc offset       297 s2
> freq     +96 delay   2802
>
>                               phc2sys[4609.778]: phc offset       197 s2
> freq     +85 delay   2799
>
>                               phc2sys[4610.778]: phc offset       137 s2
> freq     +84 delay   2790
>
>                               phc2sys[4611.779]: phc offset        63 s2
> freq     +51 delay   2811
>
>                               phc2sys[4612.779]: phc offset        49 s2
> freq     +56 delay   2754
>
>                               phc2sys[4613.780]: phc offset       -20 s2
> freq      +1 delay   2814
>
>                               phc2sys[4614.780]: phc offset        35 s2
> freq     +50 delay   2787
>
>                               phc2sys[4615.441]: phc offset        -7 s2
> freq     +19 delay   2754
>
>
> ------------------------------------------------------------------------------
> Monitor Your Dynamic Infrastructure at Any Scale With Datadog!
> Get real-time metrics from all of your servers, apps and tools
> in one place.
> SourceForge users - Click here to start your Free Trial of Datadog now!
> http://pubads.g.doubleclick.net/gampad/clk?id=241902991&iu=/4140
> _______________________________________________
> Linuxptp-users mailing list
> Lin...@li...
> https://lists.sourceforge.net/lists/listinfo/linuxptp-users
>
>

Re: [Linuxptp-users] system1: ptp4l master offset continues to climb: step_threshold 1

[Richard C. <ric...@gm...>]

On Thu, Sep 17, 2015 at 01:34:53PM -0400, Dale Smith wrote:
> I could be wrong, but isn't the step_threshold in seconds?  And -106627 is
> about  106 microseconds.  So won't it take quite a while yet before the
> threshold is reached?

Yes, you are right.  The value is in seconds.  However, I assumed that
Harold knew this, because I quoted the man page to him and setting the
threshold to 1 nanosecond makes no sense at all.

The real problem is the drifting away.  It is a sign of broken HW (or
driver), and we don't have any configuration parameter to cure that!

Thanks,
Richard

Re: [Linuxptp-users] system1: ptp4l master offset continues to climb: step_threshold 1

[Dale S. <dal...@gm...>]

Yes indeed!  I *was* going to mention that it was changing in the wrong
direction, but somehow I sent the mail before I added that.

Yep  Looks like broken HW or driver.

-Dale

On Thu, Sep 17, 2015 at 2:18 PM, Richard Cochran <ric...@gm...>
wrote:

> On Thu, Sep 17, 2015 at 01:34:53PM -0400, Dale Smith wrote:
> > I could be wrong, but isn't the step_threshold in seconds?  And -106627
> is
> > about  106 microseconds.  So won't it take quite a while yet before the
> > threshold is reached?
>
> Yes, you are right.  The value is in seconds.  However, I assumed that
> Harold knew this, because I quoted the man page to him and setting the
> threshold to 1 nanosecond makes no sense at all.
>
> The real problem is the drifting away.  It is a sign of broken HW (or
> driver), and we don't have any configuration parameter to cure that!
>
> Thanks,
> Richard
>

Re: [Linuxptp-users] system1: ptp4l master offset continues to climb: step_threshold 1

[Keller, J. E <jac...@in...>]

On Thu, 2015-09-17 at 16:04 -0400, Dale Smith wrote:
> Yes indeed!  I *was* going to mention that it was changing in the
> wrong direction, but somehow I sent the mail before I added that.
> 
> Yep  Looks like broken HW or driver.
> 
> -Dale

My gut reaction is an overflow in the frequency adjustment in the
driver, as this appears similar behavior to an issue we had on one of
the Intel drivers. If you were going to pursue this at the driver
level, I would start there.

Regards,
Jake