From: George Neville-N. <gn...@fr...> - 2013-07-10 13:40:06
|
On Jul 10, 2013, at 4:26 , Björn Dick <di...@hl...> wrote: > Hi, > > we're going to use PTPd for the synchronization of clocks on two machines with one measuring the actual power draw of the other within the Exasolvers project (http://www.sppexa.de/general-information/projects.html#EXASOLVERS). > > Within our measurement environment (sample-frequency 10kHz to 100kHz) a (more or less ;-)) 'bound' offset between clocks (of appr. 10us to 100us) is critical for the practical usage. So i'm very interested in a doubtless interpretation of the PTPd-logs! > > Am i right, that "Offset from Master" on the client-side means the "measured" offset between master and client _after_ considering the packet transmission times at the reception of a new sync-packet, i.e. the (approximated) delta with respect to an external, physical clock? > Am i also right, that the clíent's clock is ajusted at each of this sync-points by usage of the measured offset? > Yes, this is correct. Also, I've forwarded this to the ptpd list itself. Bets, George |