Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.debian.user > #263943 > unrolled thread
| Started by | gene heskett <gheskett@shentel.net> |
|---|---|
| First post | 2023-11-29 19:00 +0100 |
| Last post | 2023-11-29 22:20 +0100 |
| Articles | 20 on this page of 100 — 17 participants |
Back to article view | Back to linux.debian.user
time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-29 19:00 +0100
Re: time question, as in ntp? Dan Ritter <dsr@randomstring.org> - 2023-11-29 19:20 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-29 19:40 +0100
Re: time question, as in ntp? Dan Purgert <dan@djph.net> - 2023-11-29 19:20 +0100
Re: time question, as in ntp? Greg Wooledge <greg@wooledge.org> - 2023-11-29 20:10 +0100
Re: time question, as in ntp? Dan Purgert <dan@djph.net> - 2023-11-29 20:20 +0100
Re: time question, as in ntp? Jeffrey Walton <noloader@gmail.com> - 2023-11-30 01:30 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-29 22:00 +0100
Re: time question, as in ntp? Dan Purgert <dan@djph.net> - 2023-11-30 00:00 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-30 01:40 +0100
Re: time question, as in ntp? Dan Purgert <dan@djph.net> - 2023-11-30 11:40 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-30 11:50 +0100
Re: time question, as in ntp? Max Nikulin <manikulin@gmail.com> - 2023-11-30 03:50 +0100
Re: time question, as in ntp? Dan Purgert <dan@djph.net> - 2023-11-30 11:50 +0100
Re: time question, as in ntp? John Hasler <john@sugarbit.com> - 2023-11-29 19:30 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-29 20:00 +0100
Re: time question, as in ntp? Greg Wooledge <greg@wooledge.org> - 2023-11-29 20:30 +0100
Re: time question, as in ntp? "Andrew M.A. Cater" <amacater@einval.com> - 2023-11-29 21:00 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-30 01:10 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-30 01:00 +0100
Re: time question, as in ntp? John Hasler <john@sugarbit.com> - 2023-11-30 02:30 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-30 03:00 +0100
Re: time question, as in ntp? Charles Curley <charlescurley@charlescurley.com> - 2023-11-30 03:50 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-30 05:20 +0100
Re: time question, as in ntp? Charles Curley <charlescurley@charlescurley.com> - 2023-11-30 05:40 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-30 10:20 +0100
Re: time question, as in ntp? Karen Lewellen <klewellen@shellworld.net> - 2023-11-30 05:50 +0100
Re: time question, as in ntp? Karen Lewellen <klewellen@shellworld.net> - 2023-11-30 05:50 +0100
Re: time question, as in ntp? Dan Purgert <dan@djph.net> - 2023-11-30 11:50 +0100
Re: time question, as in ntp? Dan Purgert <dan@djph.net> - 2023-11-30 11:50 +0100
Re: time question, as in ntp? Max Nikulin <manikulin@gmail.com> - 2023-11-30 03:40 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-30 05:10 +0100
Re: time question, as in ntp? Max Nikulin <manikulin@gmail.com> - 2023-11-30 06:00 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-30 10:20 +0100
Re: time question, as in ntp? Greg Wooledge <greg@wooledge.org> - 2023-11-30 13:20 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-30 11:20 +0100
Re: time question, as in ntp? Max Nikulin <manikulin@gmail.com> - 2023-11-30 13:40 +0100
Re: time question, as in ntp? Greg Wooledge <greg@wooledge.org> - 2023-11-30 14:20 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-30 20:00 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-30 19:50 +0100
Re: time question, as in ntp? Max Nikulin <manikulin@gmail.com> - 2023-12-01 03:40 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-12-01 04:30 +0100
Re: time question, as in ntp? John Hasler <john@sugarbit.com> - 2023-12-01 05:10 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-12-01 11:10 +0100
Re: time question, as in ntp? Curt <curty@free.fr> - 2023-12-01 18:00 +0100
Re: time question, as in ntp? <tomas@tuxteam.de> - 2023-12-01 19:50 +0100
Re: time question, as in ntp? John Hasler <john@sugarbit.com> - 2023-12-01 20:30 +0100
Re: time question, as in ntp? Max Nikulin <manikulin@gmail.com> - 2023-12-01 05:20 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-12-01 11:50 +0100
Re: time question, as in ntp? Dan Purgert <dan@djph.net> - 2023-12-01 12:40 +0100
Re: time question, as in ntp? Max Nikulin <manikulin@gmail.com> - 2023-12-01 16:30 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-12-01 20:30 +0100
Re: time question, as in ntp? Greg Wooledge <greg@wooledge.org> - 2023-12-01 20:50 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-12-01 23:00 +0100
Re: time question, as in ntp? Greg Wooledge <greg@wooledge.org> - 2023-12-01 23:40 +0100
Re: time question, as in ntp? Max Nikulin <manikulin@gmail.com> - 2023-12-02 12:20 +0100
Re: time question, as in ntp? Max Nikulin <manikulin@gmail.com> - 2023-12-02 08:50 +0100
Re: time question, as in ntp? Andy Smith <andy@strugglers.net> - 2023-12-01 08:40 +0100
Re: time question, as in ntp? "Andrew M.A. Cater" <amacater@einval.com> - 2023-12-01 19:30 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-12-01 22:30 +0100
Re: time question, as in ntp? John Hasler <john@sugarbit.com> - 2023-12-02 03:40 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-12-02 11:50 +0100
Re: time question, as in ntp? John Hasler <john@sugarbit.com> - 2023-11-30 15:20 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-30 20:10 +0100
Re: time question, as in ntp? Greg Wooledge <greg@wooledge.org> - 2023-11-30 20:20 +0100
Re: time question, as in ntp? Charles Curley <charlescurley@charlescurley.com> - 2023-11-30 21:00 +0100
Re: time question, as in ntp? John Hasler <john@sugarbit.com> - 2023-11-30 21:30 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-12-01 03:10 +0100
Re: time question, as in ntp? John Hasler <john@sugarbit.com> - 2023-12-01 04:10 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-12-01 04:40 +0100
Re: time question, as in ntp? John Hasler <john@sugarbit.com> - 2023-12-01 05:20 +0100
Re: time question, as in ntp? David Wright <deblis@lionunicorn.co.uk> - 2023-12-01 06:10 +0100
Re: time question, as in ntp? Dan Purgert <dan@djph.net> - 2023-12-01 11:50 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-12-01 12:00 +0100
Re: time question, as in ntp? Dan Purgert <dan@djph.net> - 2023-12-01 12:30 +0100
Re: time question, as in ntp? Greg Wooledge <greg@wooledge.org> - 2023-12-01 14:30 +0100
Re: time question, as in ntp? Dan Purgert <dan@djph.net> - 2023-12-01 15:40 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-12-01 19:50 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-12-01 14:30 +0100
Re: time question, as in ntp? Dan Purgert <dan@djph.net> - 2023-12-01 14:50 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-12-01 19:50 +0100
Re: time question, as in ntp? Greg Wooledge <greg@wooledge.org> - 2023-12-01 04:20 +0100
Re: time question, as in ntp? Charles Curley <charlescurley@charlescurley.com> - 2023-11-30 15:20 +0100
Re: time question, as in ntp? Andy Smith <andy@strugglers.net> - 2023-11-30 17:20 +0100
Re: time question, as in ntp? Greg Wooledge <greg@wooledge.org> - 2023-11-30 19:10 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-12-01 02:40 +0100
Re: time question, as in ntp? Max Nikulin <manikulin@gmail.com> - 2023-12-01 17:30 +0100
Re: time question, as in ntp? Andy Smith <andy@strugglers.net> - 2023-12-01 21:50 +0100
Re: time question, as in ntp? Charles Curley <charlescurley@charlescurley.com> - 2023-12-02 00:10 +0100
Re: time question, as in ntp? John Hasler <john@sugarbit.com> - 2023-12-02 17:40 +0100
Re: time question, as in ntp? Max Nikulin <manikulin@gmail.com> - 2023-12-04 04:10 +0100
Re: time question, as in ntp? John Hasler <john@sugarbit.com> - 2023-12-04 05:40 +0100
Re: time question, as in ntp? Max Nikulin <manikulin@gmail.com> - 2023-12-04 11:10 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-12-04 10:50 +0100
Re: time question, as in ntp? Stefan Monnier <monnier@iro.umontreal.ca> - 2023-12-04 15:40 +0100
Re: time question, as in ntp? <tomas@tuxteam.de> - 2023-12-04 15:40 +0100
Re: time question, as in ntp? John Hasler <john@sugarbit.com> - 2023-12-04 16:00 +0100
Re: time question, as in ntp? Tom Dial <tddial@comcast.net> - 2023-12-05 09:40 +0100
Re: time question, as in ntp? Lee <ler762@gmail.com> - 2023-11-29 20:20 +0100
Re: time question, as in ntp? gene heskett <gheskett@shentel.net> - 2023-11-29 22:20 +0100
Page 5 of 5 — ← Prev page 1 2 3 4 [5]
| From | gene heskett <gheskett@shentel.net> |
|---|---|
| Date | 2023-12-01 19:50 +0100 |
| Message-ID | <HGgKl-ap85-3@gated-at.bofh.it> |
| In reply to | #264033 |
On 12/1/23 08:42, Dan Purgert wrote: > On Dec 01, 2023, gene heskett wrote: >> [lotsa snipping ... ] >> >> You claim I don't have to do anything to that printer machine, so I >> installed the ICC server here. I have done zip to the dhcpd.conf which >> looks as it it is fully disabled. Assuming I want a pool of 16 >> addresses, say from 192.168.71.100 to 192.168.71.115, what do I >> uncomment and fill in, in dhcpd.conf? > > Sent you a mail off-list with greater detail (hopefully it'll get > through). > Came in fine, than you Dan. > Simple approach --> look for the stanza beginning "subnet". It should > already be uncommented (as I recall), and probably setup for the "pretty > standard default RFC1919 range" of 192.168.1.0/24. > > > Cheers, Gene Heskett. -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author, 1940) If we desire respect for the law, we must first make the law respectable. - Louis D. Brandeis
[toc] | [prev] | [next] | [standalone]
| From | Greg Wooledge <greg@wooledge.org> |
|---|---|
| Date | 2023-12-01 04:20 +0100 |
| Message-ID | <HG2em-aeSn-1@gated-at.bofh.it> |
| In reply to | #264011 |
On Thu, Nov 30, 2023 at 08:33:47PM -0500, gene heskett wrote: > This machine has a working ntp On Thu, Nov 30, 2023 at 09:05:17PM -0500, gene heskett wrote: > let me clarify: This buster machine acting like a 3d printer does NOT have > dhcpcd installed. No trace of it in /etc Only dhcp. OK, screw it. I give up. I cannot figure out which computer you're talking about in any given sentence. You seems to switch computers every email, possibly more than once per email. If you're having a problem with your "printer", then THAT is the computer you should be talking about! But now you're using phrases like "this machine" and referring to some OTHER host that isn't the printer... GAAHH! Why would you talk about OTHER computers when you have problems with THE PRINTER?! This is impossible to understand!
[toc] | [prev] | [next] | [standalone]
| From | Charles Curley <charlescurley@charlescurley.com> |
|---|---|
| Date | 2023-11-30 15:20 +0100 |
| Message-ID | <HFQ3v-a1lN-3@gated-at.bofh.it> |
| In reply to | #263975 |
On Thu, 30 Nov 2023 11:52:04 +0700 Max Nikulin <manikulin@gmail.com> wrote: > Charles suggests to enable DHCP server on your router and I support > him. A nitpick, if I may. Apparently not all commercial routers support static addressing via DHCP. I was suggesting Gene use the ISC DHCP server, which does not have to be on your router; any Unix or Linux machine will do. Indeed, if you use the failover capability (I do), at least one instance of it will not be on the router. -- Does anybody read signatures any more? https://charlescurley.com https://charlescurley.com/blog/
[toc] | [prev] | [next] | [standalone]
| From | Andy Smith <andy@strugglers.net> |
|---|---|
| Date | 2023-11-30 17:20 +0100 |
| Message-ID | <HFRVD-a3qo-3@gated-at.bofh.it> |
| In reply to | #263948 |
Hi Gene,
On Wed, Nov 29, 2023 at 01:52:46PM -0500, gene heskett wrote:
> On 11/29/23 13:20, John Hasler wrote:
> > But first fix that address.
>
> How, John? QIDI is afraid of enabling full net access because it might
> overwrite some of their special stuff. Right now its running armbian buster,
> which is out of support. And surprise, kiauh.sh is installed, likely how
> they set the printer up in the first place. Its just a bash script but its
> magic!
No one knows what any of this stuff is because it's NOT DEBIAN and
this is DEBIAN-users. You should ask these questions on the support
venues for whatever these things are, where you will get fewer
answers that are specific to Debian.
Once you've got your networking sorted out and you are setting up an
NTP server, your next issue will be that one NTP server isn't
enough:
https://www.ntp.org/ntpfaq/ntp-s-algo-real/#532-why-should-i-have-more-than-one-clock
The reason for this is that if you have just one NTP server for your
network, and it goes bad (tells the wrong time), no one will be able
to tell.
If you have two, clients will be able to tell that one of them has
gone bad (or both!) but not which one of them.
If you have at least three, then it is possible to determine when
one of them is off.
Unless you have a dedicated time source (e.g. GPS receiver, atomic
decay source, …) then any NTP server you set up will not be
significantly better than just using the public NTP pool as all
Debian NTP clients are configured to do by default.
So basically I would say there is no point in you trying to run an
NTP server and you should just try to make sure that all your
machines having working networking and run a working NTP client.
If you DID want to run an NTP server on your network, all it would
achieve is slightly reducing the amount of traffic you send over the
Internet, however it would be entirely unsurprising for your local
NTP clients to select the remote servers as being more accurate and
send the packets over the Internet anyway. And since you need at
least least three clocks defined in every NTP client, you can't
avoid configuring at least two off-site servers. It may as well be
all three (preferably more).
But sort the networking out first.
Thanks,
Andy
--
https://bitfolk.com/ -- No-nonsense VPS hosting
[toc] | [prev] | [next] | [standalone]
| From | Greg Wooledge <greg@wooledge.org> |
|---|---|
| Date | 2023-11-30 19:10 +0100 |
| Message-ID | <HFTE5-a5lQ-1@gated-at.bofh.it> |
| In reply to | #263993 |
On Thu, Nov 30, 2023 at 04:12:48PM +0000, Andy Smith wrote: > Once you've got your networking sorted out and you are setting up an > NTP server, your next issue will be that one NTP server isn't > enough: > > https://www.ntp.org/ntpfaq/ntp-s-algo-real/#532-why-should-i-have-more-than-one-clock > > The reason for this is that if you have just one NTP server for your > network, and it goes bad (tells the wrong time), no one will be able > to tell. Well, you have to consider the actual goal. If the goal is for the printer to know the exact time, because it's going to print that time on legal documents, then your point is worth considering. But if the goal is simply to ensure that all the computers on the LAN share the *same* time, right or wrong, then having them all sync to the same, possibly drifting, time server is all that's needed. Also remember, this is just Gene's home network. It's not a major data center. A single NTP server for a home network sounds adequate to me. Ultimately it's Gene's choice to make.
[toc] | [prev] | [next] | [standalone]
| From | gene heskett <gheskett@shentel.net> |
|---|---|
| Date | 2023-12-01 02:40 +0100 |
| Message-ID | <HG0Fz-ad2J-3@gated-at.bofh.it> |
| In reply to | #263994 |
On 11/30/23 13:06, Greg Wooledge wrote: > On Thu, Nov 30, 2023 at 04:12:48PM +0000, Andy Smith wrote: >> Once you've got your networking sorted out and you are setting up an >> NTP server, your next issue will be that one NTP server isn't >> enough: >> >> https://www.ntp.org/ntpfaq/ntp-s-algo-real/#532-why-should-i-have-more-than-one-clock >> >> The reason for this is that if you have just one NTP server for your >> network, and it goes bad (tells the wrong time), no one will be able >> to tell. > > Well, you have to consider the actual goal. If the goal is for the > printer to know the exact time, because it's going to print that time > on legal documents, then your point is worth considering. > > But if the goal is simply to ensure that all the computers on the LAN > share the *same* time, right or wrong, then having them all sync to > the same, possibly drifting, time server is all that's needed. > > Also remember, this is just Gene's home network. It's not a major data > center. A single NTP server for a home network sounds adequate to me. > Ultimately it's Gene's choice to make. > This machine has a working ntp, so it should be within a millisecond or so of Boulder CO if the debian "pool" references that src. The recalcitrant armbian running the rockchip card in the printer can access this machine for ping4 or ping6 purposes. It would be adequate if for timekeeping purposes if it referenced this machine. However its just as desirable to get it "online" with the rest of my machines, which requires it get rid of the 169.254.nnn.nnn its using now. I think I've found where to do it, but not right now, the printer is busy making me a mount to solar power a you've got mail alarm, with is presently using a special 12 volt battery, a 23A, every two weeks. My checking for ntp does not find anything on that card So preliminary: 1;install on this machine the whole ntpsec group. That removed the systemd/timesyncd utility 2:create the logging dir as spec'd in /etc/ntpsec/ntp.conf. Restart /etc/init.d/ntpsec, looks like its working. I'll let it simmer a day. 3: ntp.conf describes how to make a server but i'll have to find some man page for LetsEncrypt. Nut that means I'll have to talk to Igor and rsync their repo to get it to a place on this machine where that one can suck stuff from. df says my boot drive has /dev/sdb1 863983352 15834692 804187012 2% / used, so I ought to be able to do that. Either that, or (thinking outside the box) better yet make this machine into a NAT server to that machine as that would also enable the use of kiauh to keep the printer AND its OS updated. Is there a ready made package to do that NAT-ing? Or can the modernized iptables do that?. Thanks all. Cheers, Gene Heskett. -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author, 1940) If we desire respect for the law, we must first make the law respectable. - Louis D. Brandeis
[toc] | [prev] | [next] | [standalone]
| From | Max Nikulin <manikulin@gmail.com> |
|---|---|
| Date | 2023-12-01 17:30 +0100 |
| Message-ID | <HGeyS-anWH-17@gated-at.bofh.it> |
| In reply to | #263993 |
On 30/11/2023 23:12, Andy Smith wrote: > Unless you have a dedicated time source (e.g. GPS receiver, atomic > decay source, …) A nitpick. I am puzzled by the word "decay" in this context. Electron transition between energy states in atomic clocks is not decay. Nuclear decay is hardly related to clocks as well. Observation of photon frequency change due to propagation in gravitational field of the Earth was a really impressive experiment. However radioactive decay was just a means to get excited nuclei. Emission and absorption of gamma rays is still a transition between energy states, not a decay. Anyway I am unaware of *clocks* based on Mössbauer effect. Despite extreme precision in respect to frequency change, difficulties with implementation make it impractical. As to a GPS receiver, it should be doable and 169.254.x.y addresses will not be an issue any more. Be careful with cables when connecting it however: https://www.wired.com/2012/02/neutrinos-faulty-cable/
[toc] | [prev] | [next] | [standalone]
| From | Andy Smith <andy@strugglers.net> |
|---|---|
| Date | 2023-12-01 21:50 +0100 |
| Message-ID | <HGiCt-aqey-1@gated-at.bofh.it> |
| In reply to | #264044 |
Hello, On Fri, Dec 01, 2023 at 11:21:05PM +0700, Max Nikulin wrote: > On 30/11/2023 23:12, Andy Smith wrote: > > Unless you have a dedicated time source (e.g. GPS receiver, atomic > > decay source, …) > > A nitpick. I am puzzled by the word "decay" in this context. Electron > transition between energy states in atomic clocks is not decay. I'm sure you're right; I don't really know anything about atomic clocks other than that they exist. > As to a GPS receiver, it should be doable and 169.254.x.y addresses will not > be an issue any more. Sure, all of these things are readily available to buy online and compatible with Linux. It's just so far down the list of what Gene should probably try first that it's kind of comical to have the conversation, though! The ntppool.org forum is a good place for amateur builders of "real" clocks attached to Linux/BSD. Thanks, Andy -- https://bitfolk.com/ -- No-nonsense VPS hosting
[toc] | [prev] | [next] | [standalone]
| From | Charles Curley <charlescurley@charlescurley.com> |
|---|---|
| Date | 2023-12-02 00:10 +0100 |
| Message-ID | <HGkNX-arIn-3@gated-at.bofh.it> |
| In reply to | #264044 |
On Fri, 1 Dec 2023 23:21:05 +0700 Max Nikulin <manikulin@gmail.com> wrote: > As to a GPS receiver, it should be doable and 169.254.x.y addresses > will not be an issue any more. Be careful with cables when connecting > it however: https://www.wired.com/2012/02/neutrinos-faulty-cable/ And there is plenty of expertise on the gpsd email list. But I would start with the web site, including the how-to on setting up a time server. https://gpsd.io https://gpsd.io/gpsd-time-service-howto.html -- Does anybody read signatures any more? https://charlescurley.com https://charlescurley.com/blog/
[toc] | [prev] | [next] | [standalone]
| From | John Hasler <john@sugarbit.com> |
|---|---|
| Date | 2023-12-02 17:40 +0100 |
| Message-ID | <HGBc5-aBoK-1@gated-at.bofh.it> |
| In reply to | #264072 |
Max Nikulin wrote: > As to a GPS receiver, it should be doable and 169.254.x.y addresses > will not be an issue any more. Be careful with cables when connecting > it however: https://www.wired.com/2012/02/neutrinos-faulty-cable/ CNC machines don't need accurate time. They need precise internal synchronization but that isn't related to the system clocks. The default NTP configuration in most Linux distributions will take care of the system clocks if they have access to the Internet. If not run an NTP server on one machine. -- John Hasler john@sugarbit.com Elmwood, WI USA
[toc] | [prev] | [next] | [standalone]
| From | Max Nikulin <manikulin@gmail.com> |
|---|---|
| Date | 2023-12-04 04:10 +0100 |
| Message-ID | <HH7vj-aViL-1@gated-at.bofh.it> |
| In reply to | #264109 |
On 02/12/2023 23:39, John Hasler wrote: > Max Nikulin wrote: >> As to a GPS receiver, it should be doable and 169.254.x.y addresses >> will not be an issue any more. Be careful with cables when connecting >> it however: https://www.wired.com/2012/02/neutrinos-faulty-cable/ > > CNC machines don't need accurate time. They need precise internal > synchronization but that isn't related to the system clocks. The > default NTP configuration in most Linux distributions will take care of > the system clocks if they have access to the Internet. I was kidding. However "access to the Internet" is the real trouble in this particular case. It seems, the vendor of the 3d printer believes that users must connect the device to a network having a DHCP server. ifupdown is definitely broken in armbian, perhaps armbian way to setup network is broken on this device, but Gene is not going to debug it. His stance that it is NetworkManager that breaks his network. In addition, he is strongly against DHCP believing that it will make his network unstable. From my point of view, it should be possible to put a file with mapping of mac addresses to desired IPs and names to his dd-wrt router. I expect that dnsmasq is running or can be installed there. Dnsmasq as a DHCP server on the router should be better than maintaining hosts files on each machine. The current state is that the 3d printer has only a 169.254.x.y link-local address configured as a fallback.
[toc] | [prev] | [next] | [standalone]
| From | John Hasler <john@sugarbit.com> |
|---|---|
| Date | 2023-12-04 05:40 +0100 |
| Message-ID | <HH8Up-aWvI-1@gated-at.bofh.it> |
| In reply to | #264181 |
Max Nikulin wrote: > From my point of view, it should be possible to put a file with > mapping of mac addresses to desired IPs and names to his dd-wrt > router. I expect that dnsmasq is running or can be installed > there. Dnsmasq as a DHCP server on the router should be better than > maintaining hosts files on each machine. dnsmasq will get the hostnames from the machines and put them in its dns. There is really no need to manually enter any MAC addresses or IPs anywhere. I suspect that had the machine been plugged into a network equipped with a properly configured DHCP server it would just work. -- John Hasler john@sugarbit.com Elmwood, WI USA
[toc] | [prev] | [next] | [standalone]
| From | Max Nikulin <manikulin@gmail.com> |
|---|---|
| Date | 2023-12-04 11:10 +0100 |
| Message-ID | <HHe3M-b2X4-9@gated-at.bofh.it> |
| In reply to | #264185 |
On 04/12/2023 11:38, John Hasler wrote: > Max Nikulin wrote: >> From my point of view, it should be possible to put a file with >> mapping of mac addresses to desired IPs and names to his dd-wrt >> router. I expect that dnsmasq is running or can be installed >> there. Dnsmasq as a DHCP server on the router should be better than >> maintaining hosts files on each machine. > > dnsmasq will get the hostnames from the machines and put them in its > dns. There is really no need to manually enter any MAC addresses or IPs > anywhere. IP addresses may be added to some scripts or config files already. My point is that dnsmasq allows to fix IP addresses obtained by clients through DHCP. It is first step to convince that a network managed using a DHCP server is not unstable. > I suspect that had the machine been plugged into a network equipped with > a properly configured DHCP server it would just work. Sure, but if DHCP server is not running on the router then it will be more tricky to debug next issue. Answering to questions people make enough assumptions and in home networks usually routers play this role.
[toc] | [prev] | [next] | [standalone]
| From | gene heskett <gheskett@shentel.net> |
|---|---|
| Date | 2023-12-04 10:50 +0100 |
| Message-ID | <HHdKp-b2ut-3@gated-at.bofh.it> |
| In reply to | #264181 |
On 12/3/23 22:05, Max Nikulin wrote: > On 02/12/2023 23:39, John Hasler wrote: >> Max Nikulin wrote: >>> As to a GPS receiver, it should be doable and 169.254.x.y addresses >>> will not be an issue any more. Be careful with cables when connecting >>> it however: https://www.wired.com/2012/02/neutrinos-faulty-cable/ >> >> CNC machines don't need accurate time. They need precise internal >> synchronization but that isn't related to the system clocks. The >> default NTP configuration in most Linux distributions will take care of >> the system clocks if they have access to the Internet. > > I was kidding. However "access to the Internet" is the real trouble in > this particular case. It seems, the vendor of the 3d printer believes > that users must connect the device to a network having a DHCP server. > ifupdown is definitely broken in armbian, perhaps armbian way to setup > network is broken on this device, but Gene is not going to debug it. His > stance that it is NetworkManager that breaks his network. 1. I'll retract that, while there seem to be networkmanager leftovers here and there, there is no nm present on the printer NOW that I can find. And I haven't removed it. 2. I have installed the ISC dhcpd-server on this machine, its working as expected, the printer is now inside my subnet of the 192.168.71 block. 3. I've also setup ntpsec as a server on this machine, and have the printers chrony synching to this machine but the chrony on the printer is stuck in PST, exactly 4 hours behind this machine regardless of the setting in /etc/timezone. > In addition, he is strongly against DHCP believing that it will make his > network unstable. See above, its installed and working, I've even done a small print job to prove it works. > From my point of view, it should be possible to put a file with mapping > of mac addresses to desired IPs and names to his dd-wrt router. I expect > that dnsmasq is running or can be installed there. Dnsmasq as a DHCP > server on the router should be better than maintaining hosts files on > each machine. Two reasons, the failure of those 2 seacraates within hours of each a year ago other destroyed all records of passwd's for it, and since the rot pw was 33 chars of random numbers, I'd have to do a factory reset to a broken dd-wrt (It can't do nat) and reinstall a new dd-wrt. Its working so I won't muck with it. So I put the dhcpd-server on this machine and it worked exact as Dan said it would. Then I enabled ntpsec to serve and thats working to the whole world. But the chrony on the printer is stuck in the PST timezone, ignoring the contents of /etc/timezone. So the printer is 4 hours behind me here on US/Eastern or America/NewYork zone, both seem to work correctly here. So the next question is, is ntpsec serving my time, or utc. This hdware clock is supposedly set to UTC, but what is ntpsec serving? It s/b serving UTC IMO. But I'm in the dark here, haven't had to fool with this in the last 24 years. > The current state is that the 3d printer has only a 169.254.x.y > link-local address configured as a fallback. Which is now fixed. Thanks all. Cheers, Gene Heskett. -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author, 1940) If we desire respect for the law, we must first make the law respectable. - Louis D. Brandeis
[toc] | [prev] | [next] | [standalone]
| From | Stefan Monnier <monnier@iro.umontreal.ca> |
|---|---|
| Date | 2023-12-04 15:40 +0100 |
| Message-ID | <HHih3-b5WI-7@gated-at.bofh.it> |
| In reply to | #264196 |
> So I put the dhcpd-server on this machine and it worked exact as Dan said it
> would. Then I enabled ntpsec to serve and thats working to the whole world.
> But the chrony on the printer is stuck in the PST timezone, ignoring the
> contents of /etc/timezone.
In Unix/Linux/Posix, time is counted in "seconds since the epoch" and is
timezone-agnostic. The timezone is only used when converting to/from
this number of seconds from/to the usual "year/month/day/hour/minutes/...".
> So the printer is 4 hours behind me here on
> US/Eastern or America/NewYork zone, both seem to work correctly here.
"The printer" is probably not 4 hours behind anything. Presumably it's
just some part of the printer's software which chooses to use PST when
*displaying* time information.
> So the next question is, is ntpsec serving my time, or utc.
It's serving *the* time :-)
Stefan
[toc] | [prev] | [next] | [standalone]
| From | <tomas@tuxteam.de> |
|---|---|
| Date | 2023-12-04 15:40 +0100 |
| Message-ID | <HHih4-b5WI-13@gated-at.bofh.it> |
| In reply to | #264216 |
[Multipart message — attachments visible in raw view] — view raw
On Mon, Dec 04, 2023 at 09:30:14AM -0500, Stefan Monnier wrote: [...] > It's serving *the* time :-) Well put :-) Cheers -- t
[toc] | [prev] | [next] | [standalone]
| From | John Hasler <john@sugarbit.com> |
|---|---|
| Date | 2023-12-04 16:00 +0100 |
| Message-ID | <HHiAp-b66c-5@gated-at.bofh.it> |
| In reply to | #264196 |
Gene writes:
> I've also setup ntpsec as a server on this machine, and have the
> printers chrony synching to this machine but the chrony on the printer
> is stuck in PST, exactly 4 hours behind this machine regardless of the
> setting in /etc/timezone.
Chrony only does UTC.
chronyc tracking
will tell you what time Chrony thinks it is.
--
John Hasler
john@sugarbit.com
Elmwood, WI USA
[toc] | [prev] | [next] | [standalone]
| From | Tom Dial <tddial@comcast.net> |
|---|---|
| Date | 2023-12-05 09:40 +0100 |
| Message-ID | <HHz8d-blme-1@gated-at.bofh.it> |
| In reply to | #264196 |
On 12/4/23 02:43, gene heskett wrote: > > So the next question is, is ntpsec serving my time, or utc. This hdware clock is supposedly set to UTC, but what is ntpsec serving? It s/b serving UTC IMO. But I'm in the dark here, haven't had to fool with this in the last 24 years. I'm pretty sure your NTP server will be dealing UTC time; mine, running on a GPS, does, and all the the machines on the net stay in lockstep to within a few dozen microseconds most of the time. I think you have to figure out how to tell your printer driver to set its timezone to Eastern; it's not Debian (or Red Hat or HP-UX or Solaris) and I don't have any idea. If it were debian(-derived) you could say "dpkg-reconfigure tzdata" - might be worth a try. Or maybe consult the vendor. Regards, Tom Dial > >> The current state is that the 3d printer has only a 169.254.x.y link-local address configured as a fallback. > > Which is now fixed. > > Thanks all. > > Cheers, Gene Heskett.
[toc] | [prev] | [next] | [standalone]
| From | Lee <ler762@gmail.com> |
|---|---|
| Date | 2023-11-29 20:20 +0100 |
| Message-ID | <HFygh-9JIP-1@gated-at.bofh.it> |
| In reply to | #263943 |
On Wed, Nov 29, 2023 at 12:50 PM gene heskett wrote: > > Greetings all; > > I have a 3d printer, an arm64 controller running ambian buster > it has an address of 169.254.xx.xx/16 > it can ping this machine but something is killing full net access, so it > can't set its time. With a 169.254.x.x address I'm surprised it can talk to anything else on your network. Your internet router is running dd-wrt - correct? Why not enable the dhcp server software on that and serve static IP addresses to everything on your network? eg https://wiki.dd-wrt.com/wiki/index.php/Static_DHCP Your /etc/hosts files will still work and you'll stop getting 169.254.x.x addresses assigned to your machines. Regards, Lee
[toc] | [prev] | [next] | [standalone]
| From | gene heskett <gheskett@shentel.net> |
|---|---|
| Date | 2023-11-29 22:20 +0100 |
| Message-ID | <HFA8p-9LdW-1@gated-at.bofh.it> |
| In reply to | #263951 |
On 11/29/23 14:12, Lee wrote: > On Wed, Nov 29, 2023 at 12:50 PM gene heskett wrote: >> >> Greetings all; >> >> I have a 3d printer, an arm64 controller running ambian buster >> it has an address of 169.254.xx.xx/16 >> it can ping this machine but something is killing full net access, so it >> can't set its time. > > With a 169.254.x.x address I'm surprised it can talk to anything else > on your network. > > Your internet router is running dd-wrt - correct? > Why not enable the dhcp server software on that and serve static IP > addresses to everything on your network? > eg https://wiki.dd-wrt.com/wiki/index.php/Static_DHCP That would take a full factory reset and reprogram, as I lost all those pw's a year ago when 2 brand new seagate 2T spinning rust drives went tits down off line in the night in 2 nights time before I'd had a chance to reconfigure amanda for operation under bookworm,. I rebuilt the machine with all SSD's, and that is still a g-d headache almost a year later. I have repeatedly asked why it takes a minimum of 30 seconds ( once I timed it at a few seconds over 5 minutes ) to open a path to that raid10 that I own every byte of. > Your /etc/hosts files will still work and you'll stop getting > 169.254.x.x addresses assigned to your machines. Just that one. Its probably easier to turn it into a hosts file system. Off topic, but because that printer does not have ipv6 disabled, I found I could ping6 stuff on this side of the router. Too bad its at least 100 miles away from me yet. Take care & stay well everybody. I gotta go check my snail mail box. > Regards, > Lee > . Cheers, Gene Heskett. -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author, 1940) If we desire respect for the law, we must first make the law respectable. - Louis D. Brandeis
[toc] | [prev] | [standalone]
Page 5 of 5 — ← Prev page 1 2 3 4 [5]
Back to top | Article view | linux.debian.user
csiph-web