Problems with clock on 1.1-CURRENT

EJHaug ejh at eas.slu.edu
Thu Jul 29 04:35:58 PDT 2004


Matthew Dillon wrote:
> 
>     Ok.  The only thing that I can think of that makes sense is that
>     release_timer2() is being called when it shouldn't be and this is
>     putting timer 2 into the wrong mode.
> 
>     Please try this patch and tell me if it fixes the problem.
> 
I have applied that patch, will respond about it, if it changes
anything
I have two MP boxes one ntp synced externally, the DFly system
ntp synced
to it. I started keeping npt statstics several days ago.
What i have noticed over the last several days:
I have ntpd keeping peer statistics
                            status    offset, delay and RMS
jitter, all in seconds.                                 
53214 349.957 192.168.1.219 96f4 0.363391626 0.000594580
0.004002007 0.073093508
53214 413.942 192.168.1.219 96f4 0.363391626 0.000594580
0.004760444 0.090148238
53214 478.373 192.168.1.219 80f4 0.000000000 0.000000000
0.000000000 4.000000000
53214 484.371 192.168.1.219 90f4 0.081372696 0.000447840
7.937501909 0.000003815
I do not think the line with all the zeros should be happening
also
53214 2487.510 192.168.1.219 90f4 0.042431889 0.000610412
0.004398569 1593.985387583
53214 2552.509 192.168.1.219 90f4 2982.125104276 0.000587400
0.002644682 2520.310150352
53214 2617.509 192.168.1.219 90f4 2982.125104276 0.000587400
0.002775639 2254.231757412
53214 2683.498 192.168.1.219 90f4 2982.125104276 0.000587400
0.003217436 1952.220140415
53214 2748.498 192.168.1.219 90f4 2982.125104276 0.000587400
0.002961792 1593.979623496
53214 2812.497 192.168.1.219 90f4 2982.125104276 0.000587400
0.002851636 1127.112910533
53214 2877.496 192.168.1.219 96f4 2982.125104276 0.000587400
0.003152514 0.017778813
53214 5925.657 192.168.1.219 80f4 0.000000000 0.000000000
0.000000000 4.000000000
53214 5934.655 192.168.1.219 90f4 0.003103398 0.000859799
7.937501909 0.000003815
53214 6000.639 192.168.1.219 90f4 0.003103398 0.000859799
3.937997864 0.021210114
53214 6064.623 192.168.1.219 90f4 0.045215385 0.000629922
1.938110841 0.033243855
53214 6130.607 192.168.1.219 90f4 1510.505898260 0.000562300
0.938240455 1510.4816

these giant steps are disconcerting.
One can minimize the steps to under 180 seconds by
icmp echo requests to the DFly box every 180 seconds.

ie from yesterday's peer stats log
53215 388.529 192.168.1.219 90f4 0.194028474 0.000610946
0.003900112 127.860859767
53215 452.523 192.168.1.219 96f4 128.087093594 0.000618948
0.001991596 0.038904385
53215 645.628 192.168.1.219 80f4 0.000000000 0.000000000
0.000000000 4.000000000
53215 657.623 192.168.1.219 90f4 0.003975030 0.000546498
7.937501909 0.000003815
53215 723.608 192.168.1.219 90f4 0.003975030 0.000546498
3.937997864 0.020895986

curious





More information about the Bugs mailing list