On Thursday 15 December 2005 23:16, Paul W. Abrahams wrote: <major snippage>
Anyone know?
Hi Paul, It's late and I've gotta log off, so this is a "quickie"... I once wrote up a succinct description of the ntp client process for another post, but now I can't find it to copy from :-/ From my not always perfect memory: When you first set your system up as an ntp client, the script(s) accomplish the following: 1. determines the latency between the two systems, i.e. requests the time from the ntp server N times over the course of Y period, calculates the latencies and uses the average in later calculations. 2. with the average latency measured, it again requests the current time from the ntp server and adjusts the received time for latency. With the "true" time known, it then determines if the client is "fast" or "slow" and by how much. 3. it then calculates a series of small, incremental adjustments designed to gently "nudge" the client system into sync with the server without disrupting time sensitive processes and data. 4. When the above process is completed, the client is effectively synced to the server. This makes two things possible: a) the hardware clock drift can be accurately measured and, b) that drift can be compensated for in software until the hardware clock is due for it's next adjustment (cron job? reboot?) Now, I could be "all wet" but this is how I understand the overall design of the process. HTH & regards, - Carl