From: broehl@sunee.waterloo.edu (Bernie Roehl) Subject: Re: SCI: Timestamps... Date: Thu, 6 Aug 1992 13:28:54 GMT Message-ID: Organization: University of Waterloo In article <1992Aug3.202012.26129@newshost.lanl.gov> harrison@beta.lanl.gov (David A. Harrison) writes: >Back to timestamping packets...Does anyone know of a way to >synchronize the clocks of many computers to allow useful timestamping. I'm not thrilled with the idea of timestamping packets, but the problem of synchronizing remote clocks is solvable. I'm not sure how NTP does it (the Internet Network Time Protocol) but I'm sure it's described in an RFC. I suspect it's something like: Machine A sends a packet to machine B asking it what time it is; Machine B immediately responds with a packet giving the time; Machine A measures the time from when it sent its packet to when it got the response, and adds half that delay to the time reported by B (to eliminate the travel time for the packet from B to A). It's not perfect, since the time for the inquiry packet to get from A to B may not be exactly the same as that for the response packet to get back, but it'll be close. Do it 100 times in a row and take the average, and you'll get pretty good accuracy. -- Bernie Roehl, University of Waterloo Electrical Engineering Dept Mail: broehl@sunee.waterloo.edu OR broehl@sunee.UWaterloo.ca BangPath: uunet!watmath!sunee!broehl Voice: (519) 885-1211 x 2607 [work]