Re: [chrony-dev] What Linux distributions include Chrony?

[Bill Unruh <unruh@xxxxxxxxxxxxxx>]

On Mon, 23 Nov 2009, Miroslav Lichvar wrote:

> On Mon, Nov 23, 2009 at 12:45:19PM +0100, Miroslav Lichvar wrote:
> > > By the way, I noticed in the ntp discussions that claim that ntp can handle
> > > about twice as many requests per second as can chrony. Any suspicions as to
> > > why?
> >
> > I'm seeing the opposite here. ntpd seems to stop replying at certain
> > packet rate which then uses less CPU, looks like a buffer is
> > overflowing somewhere. CPU utilization is the same when the packet
> > rate is set so both reply to everything (about 12000 pps).
>
> Here is an oprofile output for chronyd:
>
> samples  %        image name               symbol name
> 878      12.5071  chronyd                  read_from_socket
> 774      11.0256  chronyd                  get_offset_correction
> 639       9.1026  chronyd                  SCH_MainLoop
> 630       8.9744  [vdso] (tgid:31244 range:0x7fff28bff000-0x7fff28c00000) [vdso] (tgid:31244 range:0x7fff28bff000-0x7fff28c00000)
> 503       7.1652  chronyd                  TMX_GetOffsetLeft
> 464       6.6097  chronyd                  transmit_packet
> 373       5.3134  chronyd                  send_packet
> 299       4.2593  chronyd                  NSR_ProcessReceive
> 254       3.6182  chronyd                  LCL_ReadCookedTime
> 235       3.3476  chronyd                  UTI_AddDoubleToTimeval
>
>
> I think it looks good. Probably the only thing we can easily improve
> is to avoid calling TMX_GetOffsetLeft if there is no slewing currently
> running.

Yes, it looks good to me as well. One can also argue that the time delivered
by chrony is better than ntp since in ntp the only information is the current
time, while chrony can correct the current time by that slew rate correction.
Ie in ntp the clock rate and the slew rate to get rid of the measured offset
are completely mixed together, while chrony keeps them separate.

Anyway a beta sounds like a good idea.

>

--
William G. Unruh   |  Canadian Institute for|     Tel: +1(604)822-3273
Physics&Astronomy  |     Advanced Research  |     Fax: +1(604)822-5324
UBC, Vancouver,BC  |   Program in Cosmology |     unruh@xxxxxxxxxxxxxx
Canada V6T 1Z1     |      and Gravity       |  www.theory.physics.ubc.ca/

---
To unsubscribe email chrony-dev-request@xxxxxxxxxxxxxxxxxxxx with "unsubscribe" in the subject.
For help email chrony-dev-request@xxxxxxxxxxxxxxxxxxxx with "help" in the subject.
Trouble?  Email listmaster@xxxxxxxxxxxxxxxxxxxx.