Re: [ntp:questions] NTP roadmap (was Re: Poul-Henning Kamp and re-write of NTP)
skillz...@gmail.com writes: On Sunday, December 7, 2014 7:25:01 PM UTC-8, Harlan Stenn wrote: A complete rewrite of the full NTP software is slated for post-4.2.8. We're expecting (the names might be changed): - tsc - Time Sync Client: a leaf-client. PHK is working on this. Are there any plans to incorporate ideas from RADclock? The PLL that PHK mentions on his site implies a feedback model. I'm wondering if any thought has been given to a feed forward design (e.g. determine frequency based on raw ticks). Or was that model considered and rejected? Not for ntp4, at this point in time. I can see doing that for ntp5, which we're already starting to think about. And to sneak it in, we've had 2 decent donations and 2 joins since yesterday - that puts us almost 10% towards our goal! If you haven't joined or donated, please do, and also please tell other folks about this. We really need the support and every bit helps. http://nwtime.org/help-ntf-join-combined-federal-campaign/ -- Harlan Stenn st...@ntp.org http://networktimefoundation.org - be a member! ___ questions mailing list questions@lists.ntp.org http://lists.ntp.org/listinfo/questions
Re: [ntp:questions] NTP roadmap (was Re: Poul-Henning Kamp and re-write of NTP)
William Unruh writes: On 2014-12-09, skillz...@gmail.com skillz...@gmail.com wrote: On Sunday, December 7, 2014 7:25:01 PM UTC-8, Harlan Stenn wrote: Rob writes: It has been discussed before that reference clock drivers should be loadable modules or even separate processes. A complete rewrite of the full NTP software is slated for post-4.2.8. We're expecting (the names might be changed): - tsc - Time Sync Client: a leaf-client. PHK is working on this. Are there any plans to incorporate ideas from RADclock? The PLL that PHK mentions on his site implies a feedback model. I'm wondering if any thought has been given to a feed forward design (e.g. determine frequency based on raw ticks). Or was that model considered and rejected? Why would one want feedforward? You surely want the system to correct the errors in the local clock, and those are not constant or predictable. Temperature, crystal aging, maybe even computer load and voltages will all affect the clock in undetermined and undeterminable ways. Bill, have you read about RADclock? It would seem to me that the key parameter is the reach of the feedback-- does one only use the current error to correct the system, or does one use a longer stretch of errors to correct the system. The latter allow one at least in part to separate the random influences from the uniform errors-- the random errors get knocked down by the averaging of the fluctuations. Exactly which is best for time control still needs more study, but as mentioned, in the comparison of ntpd (PLL) and chrony (regression) it seems as if chrony wins in its speed of response to outside influences and in its ability to control the clock to track UTC. But I guess we all have our favourites. Yes, and your/chrony's definition of best time isn't the same as the one NTP uses. If you use NTP's model, NTP keeps better time than chrony. If you use chrony's modle, it keeps better time than NTP. -- Harlan Stenn st...@ntp.org http://networktimefoundation.org - be a member! ___ questions mailing list questions@lists.ntp.org http://lists.ntp.org/listinfo/questions
Re: [ntp:questions] NTP roadmap (was Re: Poul-Henning Kamp and re-write of NTP)
On 2014-12-09, Harlan Stenn st...@ntp.org wrote: William Unruh writes: On 2014-12-09, skillz...@gmail.com skillz...@gmail.com wrote: On Sunday, December 7, 2014 7:25:01 PM UTC-8, Harlan Stenn wrote: Rob writes: It has been discussed before that reference clock drivers should be loadable modules or even separate processes. A complete rewrite of the full NTP software is slated for post-4.2.8. We're expecting (the names might be changed): - tsc - Time Sync Client: a leaf-client. PHK is working on this. Are there any plans to incorporate ideas from RADclock? The PLL that PHK mentions on his site implies a feedback model. I'm wondering if any thought has been given to a feed forward design (e.g. determine frequency based on raw ticks). Or was that model considered and rejected? Why would one want feedforward? You surely want the system to correct the errors in the local clock, and those are not constant or predictable. Temperature, crystal aging, maybe even computer load and voltages will all affect the clock in undetermined and undeterminable ways. Bill, have you read about RADclock? No. I just looked at it quickly. I am not sure I would call it feedforward, but at least it tells me what was meant by the use of the term in this context. There are aspects of chrony which could then also be called feedforward but it is at best a mixed system. I agree with the authors that it would require extensive kernel changes, and would also make getting time from the kernel slower-- you (or the kernel) has to post process the measurements much more to deliver a time it would seem to me. It would seem to me that the key parameter is the reach of the feedback-- does one only use the current error to correct the system, or does one use a longer stretch of errors to correct the system. The latter allow one at least in part to separate the random influences from the uniform errors-- the random errors get knocked down by the averaging of the fluctuations. Exactly which is best for time control still needs more study, but as mentioned, in the comparison of ntpd (PLL) and chrony (regression) it seems as if chrony wins in its speed of response to outside influences and in its ability to control the clock to track UTC. But I guess we all have our favourites. Yes, and your/chrony's definition of best time isn't the same as the one NTP uses. If you use NTP's model, NTP keeps better time than chrony. If you use chrony's modle, it keeps better time than NTP. I do not know what these models are. I call keeping good time, the comparison of the time offered by the system clock with an independent good time source. Ie, I used a gps pps to look at the time variation given by both ntpd and chrony using a local network time negotiation with another stratum 1 pps server as the source. And the offsets of the system time disciplining the system clock from a pps source. What definitin of better do you use? Would you also say that radclock delivers worse time than ntpd under ntpd's model? (Note using NTP for ntpd is a bit confusing since ntp is primarily a time exchange protocol, and has no model) ___ questions mailing list questions@lists.ntp.org http://lists.ntp.org/listinfo/questions
Re: [ntp:questions] NTP roadmap (was Re: Poul-Henning Kamp and re-write of NTP)
Marco Marongiu writes: On 08/12/14 04:00, Harlan Stenn wrote: Several more volunteers (coders and sysadmin typs) would be great, too. Can you please elaborate on what kind of help do you need from sysadmin folks, please? We have equipment in 4 locations at the moment. Folks comfortable with any of several OSes (*BSD, Debian, Ubuntu, Solaris and OpenIndiana, even Windows); ZFS (mostly FreeBSD or FreeNAS); scripting languages (perl, sh, php, ruby, ...); firewalls; LDAP; DNS (currently BIND); IPv4 and IPv6 (configuration and IP management); SSL certificate management; websites (apache for now, TWiki, Foswiki, some static HTML sites); bugzilla and RT; IRC bots; jabber; SQL (mysql and postgresql); email (a combination of postfix and exim, with some dovecot and dspam, mailman and sympa); unix backup solutions; ssh; rsync; puppet or ansible; buildbot (or equivalent); VMs (ESXi, freebsd jails, linux containers); monitoring tools; etc. I'm sure there's more - that's the list off the top of my head. I'm not expecting *anybody* to have coverage of all of these items. Being good at some of them is all we need. -- Harlan Stenn st...@ntp.org http://networktimefoundation.org - be a member! ___ questions mailing list questions@lists.ntp.org http://lists.ntp.org/listinfo/questions
Re: [ntp:questions] NTP roadmap (was Re: Poul-Henning Kamp and re-write of NTP)
On Sunday, December 7, 2014 7:25:01 PM UTC-8, Harlan Stenn wrote: Rob writes: David Woolley david@ex.djwhome.demon.invalid wrote: It starts off a bit badly by talking about the size of nptd, when, in fact, very little of ntpd is actually involved in the core algorithm. Most of it, I guess, is reference clock drivers, plus configuration and monitoring code. It has been discussed before that reference clock drivers should be loadable modules or even separate processes. A complete rewrite of the full NTP software is slated for post-4.2.8. We're expecting (the names might be changed): - tsc - Time Sync Client: a leaf-client. PHK is working on this. Are there any plans to incorporate ideas from RADclock? The PLL that PHK mentions on his site implies a feedback model. I'm wondering if any thought has been given to a feed forward design (e.g. determine frequency based on raw ticks). Or was that model considered and rejected? - an S2+ implementation that *might* also include refclocks and can be used on leaf nodes, much like the current ntpd. - an S1 implementation that *might* also include S2 support. For ongoing support of TSC and to implement the last two we're going to need more financial support than we are currently getting. Several more volunteers (coders and sysadmin typs) would be great, too. I'm about to send out another post about donating to NTF - the short story is that if NTF can raise about $7,000 before the end of December we'll be able to qualify for an additional funding source for 2015. So that's 2 reasons for folks to join or donate to NTF before the EOYear. Please help us get there... -- Harlan Stenn st...@ntp.org http://networktimefoundation.org - be a member! ___ questions mailing list questions@lists.ntp.org http://lists.ntp.org/listinfo/questions
Re: [ntp:questions] NTP roadmap (was Re: Poul-Henning Kamp and re-write of NTP)
On 2014-12-09, skillz...@gmail.com skillz...@gmail.com wrote: On Sunday, December 7, 2014 7:25:01 PM UTC-8, Harlan Stenn wrote: Rob writes: David Woolley david@ex.djwhome.demon.invalid wrote: It starts off a bit badly by talking about the size of nptd, when, in fact, very little of ntpd is actually involved in the core algorithm. Most of it, I guess, is reference clock drivers, plus configuration and monitoring code. It has been discussed before that reference clock drivers should be loadable modules or even separate processes. A complete rewrite of the full NTP software is slated for post-4.2.8. We're expecting (the names might be changed): - tsc - Time Sync Client: a leaf-client. PHK is working on this. Are there any plans to incorporate ideas from RADclock? The PLL that PHK mentions on his site implies a feedback model. I'm wondering if any thought has been given to a feed forward design (e.g. determine frequency based on raw ticks). Or was that model considered and rejected? Why would one want feedforward? You surely want the system to correct the errors in the local clock, and those are not constant or predictable. Temperature, crystal aging, maybe even computer load and voltages will all affect the clock in undetermined and undeterminable ways. It would seem to me that the key parameter is the reach of the feedback-- does one only use the current error to correct the system, or does one use a longer stretch of errors to correct the system. The latter allow one at least in part to separate the random influences from the uniform errors-- the random errors get knocked down by the averaging of the fluctuations. Exactly which is best for time control still needs more study, but as mentioned, in the comparison of ntpd (PLL) and chrony (regression) it seems as if chrony wins in its speed of response to outside influences and in its ability to control the clock to track UTC. But I guess we all have our favourites. ___ questions mailing list questions@lists.ntp.org http://lists.ntp.org/listinfo/questions
Re: [ntp:questions] NTP roadmap (was Re: Poul-Henning Kamp and re-write of NTP)
On Sun, Dec 7, 2014 at 10:00 PM, Harlan Stenn st...@ntp.org wrote: A complete rewrite of the full NTP software is slated for post-4.2.8. Is that a a rewrite independent of of LF funded work or what you hope will flow from the LF funded work plus the further development based on matching funds? That is to say, if you don't get any donations will there still be a complete rewrite? ___ questions mailing list questions@lists.ntp.org http://lists.ntp.org/listinfo/questions
Re: [ntp:questions] NTP roadmap (was Re: Poul-Henning Kamp and re-write of NTP)
Paul writes: On Sun, Dec 7, 2014 at 10:00 PM, Harlan Stenn st...@ntp.org wrote: A complete rewrite of the full NTP software is slated for post-4.2.8. Is that a a rewrite independent of of LF funded work or what you hope will flow from the LF funded work plus the further development based on matching funds? That is to say, if you don't get any donations will there still be a complete rewrite? The rewrite is scheduled. If we get funding it will be done sooner. If we don't get funding it will happen (much) later. I'll point out that we've been on a slow countdown for 4.2.8 that has gone on for *years* longer than any of us want. That is because we have had neither the volunteers nor the funds to pay for anybody to work on it. Having spent the last many years trying to drum up the level of high-quality volunteer help we need, it became clear the easier(!) battle would be to raise funds to pay for that level of dedicated effort. That is moving slowly too, but at least for a few more months' time PHK and I have some funding from LF/CII and NTF has been created. Before NTF existed there was no way to collect funds or to do much of anything else beyond the basic NTP work. That's changed now, and it's my hope and expectation that with NTF's wider scope we'll be able to make really good leverage between the projects NTF supports, and have an easier time raising funds, too. And we've had increasing revenues at NTF every year. They are still nowhere near enough to cover even one full-time developer. The funding PHK and I get from LF/CII contains no targetted funds - it is a general stipend which mostly/partially covers general NTP development for the two of us. It does not cover any of my releng efforts, nor does it cover any sysadmin or anything else that is not strictly ntp-development related. It does not cover any hosting fees, hardware repairs or upgrades, administrative costs, phone bills, travel or expenses for Standards work or technical conferences, or anything else related to NTF's mission. By definition, the stipend PHK gets is currently sufficient to cover his current level of development work on tsc. He has more time available, but there are no funds to pay for that time. I suspect (but do not know) that he is putting in some volunteer time on this as well. The stipend I get is enough to partially cover the *development* work I am doing trying to get 4.2.8 out the door. It does not cover *anything* else, and I am putting in significant volunteer time towards getting 4.2.8 finished. It doesn't cover any of my time to follow or reply to email threads or do any support tasks, user or sysadmin or anything else like that. I am not yet getting paid for anything else I am doing for NTP or NTF. This has not been long-term stable for a long time now... Please help us keep Network Time operating they way you expect and need it to perform. Please join or donate - we've already gotten one tonight, thanks, and please keep it up! -- Harlan Stenn st...@ntp.org http://networktimefoundation.org - be a member! ___ questions mailing list questions@lists.ntp.org http://lists.ntp.org/listinfo/questions
Re: [ntp:questions] NTP roadmap (was Re: Poul-Henning Kamp and re-write of NTP)
On 08/12/14 04:00, Harlan Stenn wrote: Several more volunteers (coders and sysadmin typs) would be great, too. Can you please elaborate on what kind of help do you need from sysadmin folks, please? Ciao! -- bronto ___ questions mailing list questions@lists.ntp.org http://lists.ntp.org/listinfo/questions
