Re: [ntp:questions] NTP absolute accuracy?
Unruh wrote in message news:34jgm.50898$ph1.36...@edtnps82... [] Note that chrony will give you a factor of 2 or three improvement over ntp in the errors, assuming that the roundtrip is equally split on Linux or BSD. For those without wide-bandwidth academic connections - those folks on cable or ADSL - how good is an equal split round trip assumption? David ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
David J Taylor david-tay...@blueyonder.not-this-bit.nor-this.co.uk.invalid writes: Unruh wrote in message news:34jgm.50898$ph1.36...@edtnps82... [] Note that chrony will give you a factor of 2 or three improvement over ntp in the errors, assuming that the roundtrip is equally split on Linux or BSD. For those without wide-bandwidth academic connections - those folks on cable or ADSL - how good is an equal split round trip assumption? I have a machine with Rugby reciever that also syncs (over DSL) to a machine with a GPS unit. Based on that, it looks like the asymetry is small (not more than a handful of ms, on a path of 32ms). David. ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
David J Taylor wrote: Unruh wrote in message news:34jgm.50898$ph1.36...@edtnps82... [] Note that chrony will give you a factor of 2 or three improvement over ntp in the errors, assuming that the roundtrip is equally split on Linux or BSD. For those without wide-bandwidth academic connections - those folks on cable or ADSL - how good is an equal split round trip assumption? 2Mbit adsl here seems to give me mostly maximum variation of 5-6ms and for 50% its at 0.6ms or less. Remote connections to my servers shows offset mostly following what I see here but with occasional irregular blips of from 5ms to 200ms and these don't always seem to correspond to any local variation in load or network traffic. Heavy downloads at 230kB/s sometimes have no effect but other times I see latency as monitored by my ISP shoot up from 20ms to 100ms which I believe is down to congestion or throttling in BT network. David ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
David J Taylor wrote: For those without wide-bandwidth academic connections - those folks on cable or ADSL - how good is an equal split round trip assumption? I calculated this for my case once (3840 kbit/s down and 512 kbit/s up). The difference in transmission time due only to the different bitrates was about 1.2ms in my case. This is however only the best case, because it assumes that the uplink is not saturated. If it is, and if you are able to prioritise your upstream NTP packets over all other traffic, then you need to add half of the transmission latency of the average upstream packet to the above 1.2ms. To give you some idea, assuming that the uplink is saturated due to a file upload (i.e. MTU-sized packets), this brings the difference in latency to 14.4ms (including the above 1.2ms). If you cannot control upstream priority queuing, then the upstream latency depends on the total size of your upstream queue, which may add several MTU-sized packet transmission times to the difference (in my case, 26.5ms times three plus 1.2ms). This is then really the worst case, because it assumes that the downlink is not saturated so that there is no queueing on the ISP's side. Cheers, Jan ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
David J Taylor david-tay...@blueyonder.not-this-bit.nor-this.co.uk.invalid wrote in message news:%xqgm.126$ym4...@text.news.virginmedia.com... Unruh wrote in message news:34jgm.50898$ph1.36...@edtnps82... [] Note that chrony will give you a factor of 2 or three improvement over ntp in the errors, assuming that the roundtrip is equally split on Linux or BSD. For those without wide-bandwidth academic connections - those folks on cable or ADSL - how good is an equal split round trip assumption? David Thanks, David, David and Jan. A few milliseconds is what I had expected, so if you are on a consumer line, what implications does that have for unruh's comment? Cheers, David ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
David J Taylor wrote: David J Taylor david-tay...@blueyonder.not-this-bit.nor-this.co.uk.invalid wrote in message news:%xqgm.126$ym4...@text.news.virginmedia.com... Unruh wrote in message news:34jgm.50898$ph1.36...@edtnps82... [] Note that chrony will give you a factor of 2 or three improvement over ntp in the errors, assuming that the roundtrip is equally split on Linux or BSD. For those without wide-bandwidth academic connections - those folks on cable or ADSL - how good is an equal split round trip assumption? David Thanks, David, David and Jan. A few milliseconds is what I had expected, so if you are on a consumer line, what implications does that have for unruh's comment? I'd go for equal split being as good an approximation as any. Mostly uplinks aren't loaded too much here but I've had altq running for a few years, which increased my upload from maybe 13kB/s to near 26kB/s and at same time massive latency on saturated upload link was mostly eliminated. NTP, DNS etc are also prioritised so possibly my experience isn't typical. I might leave one of desktops running chrony for an extended period with same sources as used by one of my local servers and see how they compare. David ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
Jan Ceuleers wrote: This is however only the best case, because it assumes that the uplink is not saturated. If it is, and if you are able to prioritise your I suspect most severe asymmetry cases actually happen because the downlink is saturated. That's is exacerbated by the fact that it is more difficult to get the ISP to prioritise, than to do it yourself. ADSL has a higher download speed because most users receive much more than they transmit. ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
David J Taylor writes: On consumer-grade circuits the assumption about equal round trip is unlikely to be valid. Why would you expect the asymmetry to be worse there than elsewhere? -- John Hasler jhas...@newsguy.com Dancing Horse Hill Elmwood, WI USA ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
Unruh unruh-s...@physics.ubc.ca wrote in message news:qg_gm.50009$db2.46...@edtnps83... David J Taylor david-tay...@blueyonder.not-this-bit.nor-this.co.uk.invalid writes: David J Taylor david-tay...@blueyonder.not-this-bit.nor-this.co.uk.invalid wrote in message news:%xqgm.126$ym4...@text.news.virginmedia.com... Unruh wrote in message news:34jgm.50898$ph1.36...@edtnps82... [] Note that chrony will give you a factor of 2 or three improvement over ntp in the errors, assuming that the roundtrip is equally split on Linux or BSD. For those without wide-bandwidth academic connections - those folks on cable or ADSL - how good is an equal split round trip assumption? David Thanks, David, David and Jan. A few milliseconds is what I had expected, so if you are on a consumer line, what implications does that have for unruh's comment? Cheers, David Which comment? The one I quoted: Note that chrony will give you a factor of 2 or three improvement over ntp in the errors, assuming that the roundtrip is equally split on Linux or BSD. On consumer-grade circuits the assumption about equal round trip is unlikely to be valid. I hope that David Lord can make some tests, and I await the results with interest. Cheers, David ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
David J Taylor david-tay...@blueyonder.not-this-bit.nor-this.co.uk.invalid writes: Unruh unruh-s...@physics.ubc.ca wrote in message news:qg_gm.50009$db2.46...@edtnps83... David J Taylor david-tay...@blueyonder.not-this-bit.nor-this.co.uk.invalid writes: David J Taylor david-tay...@blueyonder.not-this-bit.nor-this.co.uk.invalid wrote in message news:%xqgm.126$ym4...@text.news.virginmedia.com... Unruh wrote in message news:34jgm.50898$ph1.36...@edtnps82... [] Note that chrony will give you a factor of 2 or three improvement over ntp in the errors, assuming that the roundtrip is equally split on Linux or BSD. For those without wide-bandwidth academic connections - those folks on cable or ADSL - how good is an equal split round trip assumption? David Thanks, David, David and Jan. A few milliseconds is what I had expected, so if you are on a consumer line, what implications does that have for unruh's comment? Cheers, David Which comment? The one I quoted: Note that chrony will give you a factor of 2 or three improvement over ntp in the errors, assuming that the roundtrip is equally split on Linux or BSD. On consumer-grade circuits the assumption about equal round trip is unlikely to be valid. As I said, chrony does a linear regression on the last n offsets. If those n offsets are all displaced by 1 sec because the outbound packets all suffer a 1 sec delay while the inbound ones do not, then the absolute time will be out by 1 sec. This is true for chrony and for ntp, and there is nothing that can be done about it. If the packets occasionally suffer a 1 sec delay but usually are the same as an inbound, then ntp will probably eliminate those via the clockfilter. Chrony can also be setup to eliminate packets whose roundtrip times are x times the lowest value found for the roundtrip, where x is settable in /etc/chrony.conf. which will also eliminate those 1 second guys. (ntp's algorithm is stupid because it eliminates 80 % of the packets, even if the excess delays are say 10% of the total delay, and keeps them even if the delay is 100 times the minimum if all of the last 8 have been in that boat. chrony's algorithm uses the minimum value of the delay over long time scales, so that if random symmetric delay noise increases many valid packets can also be thrown out. ) Asymmetric delays are something that all time systems have trouble with, and at some level there is nothing that can be done about it. Given that problem, one then needs to look at chrony and ntp's use of the data that is actually collected and ask which makes the best use of that data to estimate the rate and offset of the local clock. I think that it is clear that chrony does a far better job. Whether there are situations in which it does a worse job, I do not know, but have not been able to think of any, but that may say more about me than about chrony or ntp. ntp has had the advantage of a long history of testing in a very wide variety of situations. chrony has been out there for about 10 years, but has not been as extensively tested in adverse situations. It used to be that chrony also had the disadvantage that it did not handle leap seconds. (Ie once very few years, it would suddenly find itself out by a second and have to slew the clock for the next hour or less to correct that-- if not warned by the leapsecond flag, ntp would either step the clock (with the concomitant problems for filesystem timing), quit entirely, or take many many hours to correct the problem by slewing. ie the uncorrected leapsecond was far less of a problem for chrony than for ntp) It also did not handle any refclocks. Both of these problems have now been corrected by M Lichvar in the git repository. (chrony.tuxfamily.org) The other problem is that chrony runs only on Linux or BSD. It does NOT run on Windows. That would require a pretty extensive rewriting. Curnow got started on this ( and kept chrony very modular to try to make this easy) but it is still a big task. I hope that David Lord can make some tests, and I await the results with interest. Cheers, David ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
David J Taylor writes: On consumer-grade circuits the assumption about equal round trip is unlikely to be valid. Why would you expect the asymmetry to be worse there than elsewhere? -- John Hasler It's just an impression I had - that consumers get the worst latency and most overloaded network equipment, and no direct connection to an ultra-high-speed academic network. But I would love to learn the actual figures. David ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
Unruh wrote in message news:kx%gm.50022$db2.12...@edtnps83... [] As I said, chrony does a linear regression on the last n offsets. [] Bill, I think we have seen these remarks about chrony before. What I was wanting was a yes/no answer (although numbers would be nice) to the question: - is the 2-3 times improvement in the errors affected by the assumption of equal round trip? Perhaps there is no definitive answer? Cheers, David ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
David J Taylor writes: It's just an impression I had - that consumers get the worst latency and most overloaded network equipment... Perhaps, but it doesn't follow that they suffer a great deal of asymmetry. ...and no direct connection to an ultra-high-speed academic network. Of course not, and probably no bleeding-edge switches with packet buffering, NICs with interrupt coalescing, or experimental routers, either. As far as I know round trip symmetry is not a design goal for any part of the Net. On the other hand, deep packet inspection _could_ introduce asymmetry. Seems to me that the dumber the network between me and the server the more likely I am to see symmetric latentcy. -- John Hasler jhas...@newsguy.com Dancing Horse Hill Elmwood, WI USA ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
David J Taylor wrote: Unruh unruh-s...@physics.ubc.ca wrote in message news:qg_gm.50009$db2.46...@edtnps83... David J Taylor david-tay...@blueyonder.not-this-bit.nor-this.co.uk.invalid writes: David J Taylor david-tay...@blueyonder.not-this-bit.nor-this.co.uk.invalid wrote in message news:%xqgm.126$ym4...@text.news.virginmedia.com... Unruh wrote in message news:34jgm.50898$ph1.36...@edtnps82... [] Note that chrony will give you a factor of 2 or three improvement over ntp in the errors, assuming that the roundtrip is equally split on Linux or BSD. For those without wide-bandwidth academic connections - those folks on cable or ADSL - how good is an equal split round trip assumption? David Thanks, David, David and Jan. A few milliseconds is what I had expected, so if you are on a consumer line, what implications does that have for unruh's comment? Cheers, David Which comment? The one I quoted: Note that chrony will give you a factor of 2 or three improvement over ntp in the errors, assuming that the roundtrip is equally split on Linux or BSD. On consumer-grade circuits the assumption about equal round trip is unlikely to be valid. I hope that David Lord can make some tests, and I await the results with interest. Both fileserver using ntpd and desktop using chrony had only my local ntp servers as sources so I've restarted both with three remote sources, one common source at my isp, and alternate sources from two different sites. What I'm most interested in is the effect of temperature as it's one of two periods in year that gets largest temperature variations and I've already had to restart the servers as ntpd was unable to keep up with the higher minpoll values I had set. I'm now not quite sure how best to compare stats from chrony vs ntpd. David ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
David Lord sn...@lordynet.org writes: David J Taylor wrote: Unruh unruh-s...@physics.ubc.ca wrote in message news:qg_gm.50009$db2.46...@edtnps83... David J Taylor david-tay...@blueyonder.not-this-bit.nor-this.co.uk.invalid writes: David J Taylor david-tay...@blueyonder.not-this-bit.nor-this.co.uk.invalid wrote in message news:%xqgm.126$ym4...@text.news.virginmedia.com... Unruh wrote in message news:34jgm.50898$ph1.36...@edtnps82... [] Note that chrony will give you a factor of 2 or three improvement over ntp in the errors, assuming that the roundtrip is equally split on Linux or BSD. For those without wide-bandwidth academic connections - those folks on cable or ADSL - how good is an equal split round trip assumption? David Thanks, David, David and Jan. A few milliseconds is what I had expected, so if you are on a consumer line, what implications does that have for unruh's comment? Cheers, David Which comment? The one I quoted: Note that chrony will give you a factor of 2 or three improvement over ntp in the errors, assuming that the roundtrip is equally split on Linux or BSD. On consumer-grade circuits the assumption about equal round trip is unlikely to be valid. I hope that David Lord can make some tests, and I await the results with interest. Both fileserver using ntpd and desktop using chrony had only my local ntp servers as sources so I've restarted both with three remote sources, one common source at my isp, and alternate sources from two different sites. What I'm most interested in is the effect of temperature as it's one of two periods in year that gets largest temperature variations and I've already had to restart the servers as ntpd was unable to keep up with the higher minpoll values I had set. I'm now not quite sure how best to compare stats from chrony vs ntpd. Well, the way I did it was to attach a GPS PPS to the computer and use the offset between it and the system time as a measurement of the accuracy. (the gps was NOT used as a source for eitehr ntp or chrony). David ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
David J Taylor david-tay...@blueyonder.not-this-bit.nor-this.co.uk.invalid writes: Unruh wrote in message news:kx%gm.50022$db2.12...@edtnps83... [] As I said, chrony does a linear regression on the last n offsets. [] Bill, I think we have seen these remarks about chrony before. Some have, some have not. I was just putting out there what the difference between ntp and chrony was, so people could make up their own minds. What I was wanting was a yes/no answer (although numbers would be nice) to the question: - is the 2-3 times improvement in the errors affected by the assumption of equal round trip? Perhaps there is no definitive answer? As I tried to emphasise, if the round trip is not symmetric, then neither ntp not chrony can compensate for that lack of symmetry, and the absolute time will be out. If occasionally it has an assymetric round trip, then ntp will probably elimate it, and chrony may or may not, depending on how you set up chrony, but typically chrony will be more sensitive to that asymmetry. Eb, say on exactly every 5th query, an extra .3ms is added to the return trip, and lets say that the minimum round trip is .15ms plus or minus .1 ntp will always throw out those fifth cases. but will have an error around .05 ms. chrony will not, and without them would have an error of .02ms but with them would have an error of about .05ms, just as would ntp. Ie, in this case the chrony advantage would be more or less erased by those asymmetric round trips. There is no definative answer. It depends on the situation. Cheers, David ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] NTP absolute accuracy?
Unruh wrote: David Lord sn...@lordynet.org writes: David J Taylor wrote: Unruh unruh-s...@physics.ubc.ca wrote in message news:qg_gm.50009$db2.46...@edtnps83... David J Taylor david-tay...@blueyonder.not-this-bit.nor-this.co.uk.invalid writes: David J Taylor david-tay...@blueyonder.not-this-bit.nor-this.co.uk.invalid wrote in message news:%xqgm.126$ym4...@text.news.virginmedia.com... Unruh wrote in message news:34jgm.50898$ph1.36...@edtnps82... [] Note that chrony will give you a factor of 2 or three improvement over ntp in the errors, assuming that the roundtrip is equally split on Linux or BSD. For those without wide-bandwidth academic connections - those folks on cable or ADSL - how good is an equal split round trip assumption? David Thanks, David, David and Jan. A few milliseconds is what I had expected, so if you are on a consumer line, what implications does that have for unruh's comment? Cheers, David Which comment? The one I quoted: Note that chrony will give you a factor of 2 or three improvement over ntp in the errors, assuming that the roundtrip is equally split on Linux or BSD. On consumer-grade circuits the assumption about equal round trip is unlikely to be valid. I hope that David Lord can make some tests, and I await the results with interest. Both fileserver using ntpd and desktop using chrony had only my local ntp servers as sources so I've restarted both with three remote sources, one common source at my isp, and alternate sources from two different sites. What I'm most interested in is the effect of temperature as it's one of two periods in year that gets largest temperature variations and I've already had to restart the servers as ntpd was unable to keep up with the higher minpoll values I had set. I'm now not quite sure how best to compare stats from chrony vs ntpd. Well, the way I did it was to attach a GPS PPS to the computer and use the offset between it and the system time as a measurement of the accuracy. (the gps was NOT used as a source for eitehr ntp or chrony). How do you get the time difference between your GPS and system time? At the moment there only seems to be 100us between offsets and apart from this being insignificant I wouldn't know which was nearer to ntp time. My adsl connection has also been off already for 30 minutes after I'd set this up. One of my servers that gets time from MSF and also has both test systems as sources. Both test systems also have the common source of my isp's timeserver so I'm expecting to get something meaningful from that. I might be able to get GPS to both systems but not for a while. David ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
