[ntp:questions] SIGINFO for Portable OpenNTP on Linux
I'm using OpenNTP 3.9p1 (http://www.openntpd.org/) on an RedHat 8 Linux server. The docs at http://www.openbsd.org/cgi-bin/man.cgi?query=ntpd say When ntpd receives a SIGINFO signal, it writes its peer and sensor status to syslog(3). However SIGINFO isn't defined on Linux. Has this feature been removed from the Portable version of OpenNTP or is there some other signal I can send, or some other way to get peer status (hopefully similar to ntpq -p on xntpd) to monitor convergence etc. -- RGB ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] SIGINFO for Portable OpenNTP on Linux
On Fri, Sep 4, 2009 at 9:44 AM, someone too cool to use a working email address wrote: I'm using OpenNTP 3.9p1 (http://www.openntpd.org/) on an RedHat 8 Linux server. This list (and ntp.org in general) are not related to OpenNTP. I'm afraid you'll need to ask the openntpd.org or openbsd.org folks. Good luck, Dave Hart ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] SIGINFO for Portable OpenNTP on Linux
On Sep 4, 9:44 am, RedGrittyBrick wrote: I'm using OpenNTP 3.9p1 (http://www.openntpd.org/) on an RedHat 8 Linux server. This group (which is theoretically also gatewayed to questions@lists.ntp.org) is populated mostly by people using the ntp.org reference implementation. You'll probably have more success with a OpenNTPD- or OpenBSD-focused list. If you'd like some Chrony advocacy, we have a few participants willing to oblige. Cheers, Dave Hart ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] SIGINFO for Portable OpenNTP on Linux
RedGrittyBrick redgrittybr...@spamweary.invalid wrote in message news:4aa0e178$0$2541$da0fe...@news.zen.co.uk... I'm using OpenNTP 3.9p1 (http://www.openntpd.org/) on an RedHat 8 Linux server. Oh great. Now you'll get a hundred people telling you that OpenNTP is Not Real NTP. The docs at http://www.openbsd.org/cgi-bin/man.cgi?query=ntpd say When ntpd receives a SIGINFO signal, it writes its peer and sensor status to syslog(3). Is this the man page for (reference) NTP or does it apply to OpenNTP? However SIGINFO isn't defined on Linux. Has this feature been removed from the Portable version of OpenNTP or is there some other signal I can send, or some other way to get peer status (hopefully similar to ntpq -p on xntpd) to monitor convergence etc. According to Wikipedia, SIGINFO is 'a synonym for' SIGPWR on Linux. You might try that instead. Presumably this would be somewhat easy to find in the source. Groetjes, Maarten Wiltink ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] SIGINFO for Portable OpenNTP on Linux
On Sep 4, 11:42 am, Maarten Wiltink maar...@kittensandcats.net wrote: RedGrittyBrick redgrittybr...@spamweary.invalid wrote in message news:4aa0e178$0$2541$da0fe...@news.zen.co.uk... I'm using OpenNTP 3.9p1 (http://www.openntpd.org/) on an RedHat 8 Linux server. Oh great. Now you'll get a hundred people telling you that OpenNTP is Not Real NTP. It's a knockoff ntpd whose creator chose to use the same executable name as the reference implementation. Since this group is, for better or worse, focused on that reference implementation, let me be the first of many. OpenNTPD is Not Real NTPD. I mean it's not implementing the draft NTPv4 specification. It is, on the other hand, a daemon which uses the NTP protocol. The docs athttp://www.openbsd.org/cgi-bin/man.cgi?query=ntpdsay When ntpd receives a SIGINFO signal, it writes its peer and sensor status to syslog(3). Is this the man page for (reference) NTP or does it apply to OpenNTP? Read it. It's for The ntpd daemon Any questions? The confusion is apparently intentional. OpenNTPD's creators seem to live in a world where There Can Be Only One, and to them, the One is what some refer to as OpenNTPD. In case it's not already clear, I wish the OpenNTPD creators had taken the approach of DragonflyBSD, another OS which chooses to distribute their own NTP daemon: http://leaf.dragonflybsd.org/cgi/web-man?section=8command=dntpd Thanks to inertia, I wouldn't bet on the name of OpenBSD's ntpd ever changing. But I can grumble. ;) Cheers, Dave Hart ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] SIGINFO for Portable OpenNTP on Linux
Maarten Wiltink wrote: RedGrittyBrick redgrittybr...@spamweary.invalid wrote in message news:4aa0e178$0$2541$da0fe...@news.zen.co.uk... I'm using OpenNTP 3.9p1 (http://www.openntpd.org/) on an RedHat 8 Linux server. Oh great. Now you'll get a hundred people telling you that OpenNTP is Not Real NTP. Oh dear. The docs at http://www.openbsd.org/cgi-bin/man.cgi?query=ntpd say When ntpd receives a SIGINFO signal, it writes its peer and sensor status to syslog(3). Is this the man page for (reference) NTP or does it apply to OpenNTP? It is the man page referred to by the Openntp web site. However SIGINFO isn't defined on Linux. Has this feature been removed from the Portable version of OpenNTP or is there some other signal I can send, or some other way to get peer status (hopefully similar to ntpq -p on xntpd) to monitor convergence etc. According to Wikipedia, SIGINFO is 'a synonym for' SIGPWR on Linux. You might try that instead. Thanks Presumably this would be somewhat easy to find in the source. I had looked, there's a couple of sighndlr routines that don't mention any signals other than TERM, INT and HUP. My C skills are limited so I thought it possible I'd missed something. -- RGB ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] SIGINFO for Portable OpenNTP on Linux
RedGrittyBrick wrote: My aim is simply to get timestamps in logfiles synchronised to within 100ms across a handful of servers and PCs on a LAN. I'm aware that NTP normally does far far better than this, it may be overkill for my needs. Can you use logging over the net ( udp, port 514 ) to a single host ? You get double timestamps one from the originating host, the other from the logging/receiving host. uwe ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
[ntp:questions] ntpd not talking to servers? (was Re: SIGINFO for Portable OpenNTP on Linux)
RedGrittyBrick wrote: Dave Hart wrote: On Sep 4, 9:44 am, RedGrittyBrick wrote: I'm using OpenNTP 3.9p1 (http://www.openntpd.org/) on an RedHat 8 Linux server. This group (which is theoretically also gatewayed to questions@lists.ntp.org) is populated mostly by people using the ntp.org reference implementation. ... RedHat 8 does have ntp-4.1.1a which is the latest RPM I can find for it. However this runs but has problems communicating with NTP servers. I've not been able to fathom why. Maybe I should revisit that. Here's what happens # rpm -q ntp ntp-4.1.1a-9 # grep ^[^#] /etc/ntp.conf restrict default ignore restrict 127.0.0.1 server 0.uk.pool.ntp.org server 1.uk.pool.ntp.org server 2.uk.pool.ntp.org server 3.uk.pool.ntp.org driftfile /etc/ntp/drift broadcastdelay 0.008 authenticate yes keys/etc/ntp/keys # /etc/init.d/ntpd start Starting ntpd: [ OK ] # grep ntp /var/log/messages Sep 4 15:00:40 amber ntpd: ntpd startup succeeded Sep 4 15:00:40 amber ntpd[3083]: ntpd 4.1...@1.791 Sat Aug 31 18:27:29 EDT 2002 (1) Sep 4 15:00:40 amber ntpd[3083]: precision = 13 usec Sep 4 15:00:40 amber ntpd[3083]: kernel time discipline status 0040 Sep 4 15:00:40 amber ntpd[3083]: frequency initialized 0.000 from /etc/ntp/drift # ntpq -p remote refid st t when poll reach delay offset jitter == ntp0.cis.strath 0.0.0.0 16 u- 6400.0000.000 4000.00 ntp1.exa.net.uk 0.0.0.0 16 u- 6400.0000.000 4000.00 dns1.rmplc.co.u 0.0.0.0 16 u- 6400.0000.000 4000.00 odie.tomelliott 0.0.0.0 16 u- 6400.0000.000 4000.00 # date Fri Sep 4 15:40:21 BST 2009 # ntpq -p remote refid st t when poll reach delay offset jitter == ntp0.cis.strath 0.0.0.0 16 u- 12800.0000.000 4000.00 ntp1.exa.net.uk 0.0.0.0 16 u- 12800.0000.000 4000.00 dns1.rmplc.co.u 0.0.0.0 16 u- 12800.0000.000 4000.00 odie.tomelliott 0.0.0.0 16 u- 12800.0000.000 4000.00 # ntpdate -q ntp1.exa.net.uk server 82.219.4.30, stratum 2, offset -0.061504, delay 0.07434 4 Sep 15:40:57 ntpdate[3311]: adjust time server 82.219.4.30 offset -0.061504 sec So, no contact with NTP servers after 40 mins of running, even though ntpdate can query those same servers quite happily. Any ideas how I can find out what's wrong? -- RGB ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] ntpd not talking to servers? (was Re: SIGINFO for Portable OpenNTP on Linux)
Ronan Flood wrote: RedGrittyBrick redgrittybr...@spamweary.invalid wrote: So, no contact with NTP servers after 40 mins of running, even though ntpdate can query those same servers quite happily. ntpdate -q uses a high local port, ntpd uses local port 123. Oh, OK. If you stop ntpd, does ntpdate, without -q or -u or -d, work? # ntpdate 4 Sep 16:35:22 ntpdate[3523]: no servers can be used, exiting # /etc/init.d/ntpd stop Shutting down ntpd:[ OK ] # /etc/init.d/ntpd status ntpd is stopped # ntpdate 4 Sep 16:35:49 ntpdate[3543]: no servers can be used, exiting If not, perhaps a firewall issue. The openntpd daemon seemed to be able to talk to servers, here's some syslog records from it: Sep 4 09:47:27 amber ntpd[1261]: ntp engine ready Sep 4 09:47:44 amber ntpd[1261]: peer 212.13.195.3 now valid Sep 4 09:47:45 amber ntpd[1261]: peer 93.89.85.202 now valid Sep 4 09:47:49 amber ntpd[1261]: peer 194.238.48.3 now valid Sep 4 09:47:50 amber ntpd[1261]: peer 91.213.50.15 now valid Sep 4 09:47:51 amber ntpd[1261]: peer 82.219.4.30 now valid Sep 4 10:03:52 amber ntpd[1261]: clock is now synced Sep 4 10:16:51 amber ntpd[1261]: peer 93.89.85.202 now invalid Sep 4 10:28:20 amber ntpd[1261]: peer 93.89.85.202 now valid Sep 4 10:31:24 amber ntpd[1261]: clock is now unsynced Sep 4 10:50:52 amber ntpd[1261]: clock is now synced Sep 4 11:08:28 amber ntpd[1260]: adjusting local clock by -0.135589s Sep 4 11:32:22 amber ntpd[1261]: clock is now unsynced Sep 4 11:40:02 amber ntpd[1261]: peer 194.238.48.3 now invalid Sep 4 11:40:57 amber ntpd[1261]: peer 93.89.85.202 now invalid Sep 4 11:42:30 amber ntpd[1261]: clock is now synced Sep 4 11:53:33 amber ntpd[1261]: peer 194.238.48.3 now valid Sep 4 11:53:53 amber ntpd[1261]: peer 93.89.85.202 now valid Sep 4 12:28:08 amber ntpd[1260]: adjusting local clock by -0.191927s Sep 4 12:37:32 amber ntpd[1261]: clock is now unsynced Sep 4 12:57:26 amber ntpd[1261]: clock is now synced Sep 4 13:37:16 amber ntpd[1260]: adjusting local clock by -0.190155s Sep 4 13:37:16 amber ntpd[1261]: clock is now unsynced Sep 4 13:44:14 amber ntpd[1261]: clock is now synced Sep 4 13:47:21 amber ntpd[1261]: peer 194.238.48.3 now invalid Sep 4 13:58:46 amber ntpd[1261]: peer 194.238.48.3 now valid Sep 4 14:05:51 amber ntpd[1261]: peer 212.13.195.3 now invalid Sep 4 14:19:06 amber ntpd[1261]: peer 212.13.195.3 now valid Sep 4 14:31:32 amber ntpd[1260]: adjusting local clock by -0.137812s Sep 4 14:31:32 amber ntpd[1261]: clock is now unsynced Sep 4 14:44:28 amber ntpd[1260]: adjusting local clock by -0.154506s Sep 4 14:44:28 amber ntpd[1261]: clock is now synced Sep 4 14:53:47 amber ntpd[1260]: adjusting local clock by -0.170122s Sep 4 14:58:06 amber ntpd[1261]: clock is now unsynced Sep 4 15:00:12 amber ntpd[1261]: ntp engine exiting Sep 4 15:00:12 amber ntpd[1260]: Terminating I can't say I follow what it is doing, but I do get the impression it is able to talk to NTP servers. A non-Linux server on the same LAN can use ntpd OK. A Windows XP client can get time from 0.uk.pool.ntp.org using UDP on port 123. A sniffer (Wireshark) shows NTP version 3 packets. Here's what tcpdump has to say on the non-working box when I start ntpd # tcpdump -s0 port 123 tcpdump: listening on eth0 17:03:10.167926 amber.my.domain.ntp razor.peloris.co.uk.ntp: v4 client strat 0 poll 6 prec -17 (DF) [tos 0x10] 17:03:10.231563 razor.peloris.co.uk.ntp amber.my.domain.ntp: v4 server strat 2 poll 6 prec -20 (DF) 17:03:19.167366 amber.my.domain.ntp ntp0.lon.bitfolk.com.ntp: v4 client strat 0 poll 6 prec -17 (DF) [tos 0x10] 17:03:19.223485 ntp0.lon.bitfolk.com.ntp amber.my.domain.ntp: v4 server strat 3 poll 6 prec -20 (DF) 17:03:20.167292 amber.my.domain.ntp ntp1.arse.org.ntp: v4 client strat 0 poll 6 prec -17 (DF) [tos 0x10] 17:03:23.167106 amber.my.domain.ntp ntp1.exa.net.uk.ntp: v4 client strat 0 poll 6 prec -17 (DF) [tos 0x10] 17:03:23.215801 ntp1.exa.net.uk.ntp amber.my.domain.ntp: v4 server strat 2 poll 6 prec -20 (DF) # ntpq -p remote refid st t when poll reach delay offset jitter == ntp1.arse.org 0.0.0.0 16 u- 6400.0000.000 4000.00 ntp0.lon.bitfol 0.0.0.0 16 u- 6400.0000.000 4000.00 ntp1.exa.net.uk 0.0.0.0 16 u- 6400.0000.000 4000.00 razor.peloris.c 0.0.0.0 16 u- 6400.0000.000 4000.00 Odd eh? There's network traffic. Why doesn't it show up in `ntpq -p`? -- RGB ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] ntpd not talking to servers? (was Re: SIGINFO for Portable OpenNTP on Linux)
RedGrittyBrick redgrittybr...@spamweary.invalid wrote: # ntpdate 4 Sep 16:35:49 ntpdate[3543]: no servers can be used, exiting You have to give it a server: ntpdate ntp1.exa.net.uk or whatever. However thanks for the tcpdump, which does indeed show traffic flowing. Looking again at your ntp.conf, you've got restrict default ignore which will be the problem here. You have to open the restrictions for the servers you're trying to use, but that might be complicated by the fact that you're using the pool, with multiple A records for each entry if your DNS resolver is rotating entries. Try restrict default nomodify nopeer notrap instead of ignore and see how that goes. -- Ronan Flood use...@umbral.org.uk ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
[ntp:questions] Need patch for [BUG 452]
Hi, Please, could anyone point me the patch for this bug: [Bug 452] Do not report kernel PLL/FLL flips. Thanks, Anand ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] SIGINFO for Portable OpenNTP on Linux
Dave Hart daveh...@gmail.com writes: On Sep 4, 1:55=A0pm, RedGrittyBrick wrote: RedHat 8 does have ntp-4.1.1a which is the latest RPM I can find for it. However this runs but has problems communicating with NTP servers. I've =A0 not been able to fathom why. Maybe I should revisit that. I wouldn't bother. That version came out in 2002 [1], any problems are not going to garner a lot of interest from the developers of ntp unless they also occur on the current ntp-stable (4.2.4p7) or recent ntp-dev (4.2.5) as well. I would download the latest ntp.org source but my C skills are few and I thought there might be dependency issues on such an old OS. Any advice/suggestions welcome. Now you're talking my language. You don't need to have any C skills to use the ntp.org source tarballs. Grab 4.2.4p7 or 4.2.5p209 from http://www.ntp.org/downloads.html, then: tar xzf ntp*.gz cd ntp* ./configure --with-lineeditlibs=3Dreadline make Well, you had better make sure that you have a variety of developement packages installed. Eg, gcc, make, libc-devel, ... You should then either: sudo make install -or- su make install If you have any problems, we'll help you through them. The ntp.org distribution works on some very ancient OSes, yours is still relatively young by our standards :) Cheers, Dave Hart [1] http://archive.ntp.org/ntp4/ntp-4.1/ ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] SIGINFO for Portable OpenNTP on Linux
Dave Hart wrote: On Sep 4, 1:55 pm, RedGrittyBrick wrote: RedHat 8 does have ntp-4.1.1a which is the latest RPM I can find for it. However this runs but has problems communicating with NTP servers. I've not been able to fathom why. Maybe I should revisit that. I wouldn't bother. That version came out in 2002 [1], any problems are not going to garner a lot of interest from the developers of ntp unless they also occur on the current ntp-stable (4.2.4p7) or recent ntp-dev (4.2.5) as well. I would download the latest ntp.org source but my C skills are few and I thought there might be dependency issues on such an old OS. Any advice/suggestions welcome. Now you're talking my language. You don't need to have any C skills to use the ntp.org source tarballs. Grab 4.2.4p7 or 4.2.5p209 from http://www.ntp.org/downloads.html, then: tar xzf ntp*.gz cd ntp* ./configure --with-lineeditlibs=readline make You should then either: sudo make install -or- su make install If you have any problems, we'll help you through them. The ntp.org distribution works on some very ancient OSes, yours is still relatively young by our standards :) I downloaded, configured, made, installed and ran 4.2.4p7. Same problem. I changed the ntp.conf as Ronan suggested, restarted ntpd and all is well. I'm happy. Thanks Dave. -- RGB ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] ntpd not talking to servers? (was Re: SIGINFO for Portable OpenNTP on Linux)
Ronan Flood wrote: RedGrittyBrick redgrittybr...@spamweary.invalid wrote: # ntpdate 4 Sep 16:35:49 ntpdate[3543]: no servers can be used, exiting You have to give it a server: ntpdate ntp1.exa.net.uk or whatever. However thanks for the tcpdump, which does indeed show traffic flowing. Looking again at your ntp.conf, you've got restrict default ignore which will be the problem here. You have to open the restrictions for the servers you're trying to use, but that might be complicated by the fact that you're using the pool, with multiple A records for each entry if your DNS resolver is rotating entries. Try restrict default nomodify nopeer notrap instead of ignore and see how that goes. That fixed it, thanks! -- RGB ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] ntpd not talking to servers? (was Re: SIGINFO for Portable OpenNTP on Linux)
On 2009-09-04, RedGrittyBrick redgrittybr...@spamweary.invalid wrote: # grep ^[^#] /etc/ntp.conf restrict default ignore restrict 127.0.0.1 server 0.uk.pool.ntp.org server 1.uk.pool.ntp.org server 2.uk.pool.ntp.org server 3.uk.pool.ntp.org The 'restrict default ignore' line is your problem. It tells ntpd to ignore _all_ NTP packets from _anywhere_. When 'restrict default ignore' is used one must explicitly allow packets from authorized addresses (which can be servers or clients). Unfortunately, this won't work with the pool or any other host name which can resolve to multiple IP addresses. A more sensible set of restriction lines is: restrict default nomodify nopeer notrap noquery restrict 127.0.0.1 More information about restrictions is available at http://support.ntp.org/Support/AccessRestrctions. driftfile /etc/ntp/drift These lines aren't doing anything useful: broadcastdelay 0.008 authenticate yes keys/etc/ntp/keys -- Steve Kostecke koste...@ntp.org NTP Public Services Project - http://support.ntp.org/ ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] ntpd not talking to servers? (was Re: SIGINFO for Portable OpenNTP on Linux)
On 2009-09-04, RedGrittyBrick redgrittybr...@spamweary.invalid wrote: # ntpdate 4 Sep 16:35:22 ntpdate[3523]: no servers can be used, exiting That won't work. You have to pass the server name(s) on the command line. e.g. 'ntpdate pool.ntp.org' -- Steve Kostecke koste...@ntp.org NTP Public Services Project - http://support.ntp.org/ ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] SIGINFO for Portable OpenNTP on Linux
On 2009-09-04, Dave Hart daveh...@gmail.com wrote: On Sep 4, 9:44 am, RedGrittyBrick wrote: I'm using OpenNTP 3.9p1 (http://www.openntpd.org/) on an RedHat 8 Linux server. This group (which is theoretically also gatewayed to questions@lists.ntp.org) See https://lists.ntp.org/pipermail/questions/2009-September/thread.html -- Steve Kostecke koste...@ntp.org NTP Public Services Project - http://support.ntp.org/ ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] SIGINFO for Portable OpenNTP on Linux
On 2009-09-04, RedGrittyBrick redgrittybr...@spamweary.invalid wrote: Dave Hart wrote: This group ... is populated mostly by people using the ntp.org reference implementation. I didn't realise this was a controversial point, sorry for any annoyance caused. It's not a matter of annoyance. It's the simple fact that, AFAIK, no one here has any real experience (at least, that they'll admit to) with that other implementation. RedHat 8 does have ntp-4.1.1a which is the latest RPM I can find for it. However this runs but has problems communicating with NTP servers. I've not been able to fathom why. Maybe I should revisit that. Yes, you should. -- Steve Kostecke koste...@ntp.org NTP Public Services Project - http://support.ntp.org/ ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] SIGINFO for Portable OpenNTP on Linux
On Sep 4, 3:51 pm, Steve Kostecke wrote: On 2009-09-04, Dave Hart wrote: This group (which is theoretically also gatewayed to questi...@lists.ntp.org) Seehttps://lists.ntp.org/pipermail/questions/2009-September/thread.html It's a one-way gateway in practice. Newsgroup postings make it to questions@ but quesstions@ traffic never makes it to the newsgroup. This has been a problem all summer and has as has been noted recently by Martin Burnicki and David J Taylor. In this very thread, you will find my first reply was to questions@lists.ntp.org, and has not made it to the newsgroup. [1] I switched over to posting to the newsgroup, since otherwise the OP would likely never see my messages. Cheers, Dave Hart [1] Date: Fri, 4 Sep 2009 11:37:51 + Message-ID: 85d954180909040437i5eb57d8dt489ebe0c9ad6f...@mail.gmail.com http://lists.ntp.org/pipermail/questions/2009-September/024170.html ^ note lack of s means no dire browser warnings... ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] ntpd not talking to servers?
Your problem is the 'restrict' line in your ntp.conf file. Please see: http://support.ntp.org/Support/ConfiguringNTP H ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] Need patch for [BUG 452]
In article 232192a8-6957-42ea-ad40-96dc60393...@y10g2000prg.googlegroups.com, Mysore mr.anandaku...@gmail.com writes: Mysore Hi, Please, could anyone point me the patch for this bug: [Bug 452] Mysore Do not report kernel PLL/FLL flips. Sure: - look at the bug report to see when it was fixed and if the fix is in -stable (and therefore -dev) or -dev only - visit http://ntp.bkbits.net, pick ntp-stable (if the fix is there) or ntp-dev (otherwise) - select the appropriate date range - search for 452 - visit the changeset - see the patches -- Harlan Stenn st...@ntp.org http://ntpforum.isc.org - be a member! ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] SIGINFO for Portable OpenNTP on Linux
RedGrittyBrick wrote: I'm using OpenNTP 3.9p1 (http://www.openntpd.org/) on an RedHat 8 Linux server. The docs at http://www.openbsd.org/cgi-bin/man.cgi?query=ntpd say When ntpd receives a SIGINFO signal, it writes its peer and sensor status to syslog(3). However SIGINFO isn't defined on Linux. Has this feature been removed from the Portable version of OpenNTP or is there some other signal I can send, or some other way to get peer status (hopefully similar to ntpq -p on xntpd) to monitor convergence etc. Open mouth, insert foot, chew vigorously! OpenNTP is do it yourself project by some group whose name and affiliations I can't remember. It *does* use the NTP protocol but there the connection ends! It's possible that members of the responsible group monitor this news group but I wouldn't count on it. ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
Re: [ntp:questions] ntpd not talking to servers? (was Re: SIGINFO for Portable OpenNTP on Linux)
RedGrittyBrick wrote: RedGrittyBrick wrote: Dave Hart wrote: On Sep 4, 9:44 am, RedGrittyBrick wrote: I'm using OpenNTP 3.9p1 (http://www.openntpd.org/) on an RedHat 8 Linux server. This group (which is theoretically also gatewayed to questions@lists.ntp.org) is populated mostly by people using the ntp.org reference implementation. ... RedHat 8 does have ntp-4.1.1a which is the latest RPM I can find for it. However this runs but has problems communicating with NTP servers. I've not been able to fathom why. Maybe I should revisit that. Here's what happens # rpm -q ntp ntp-4.1.1a-9 # grep ^[^#] /etc/ntp.conf restrict default ignore restrict 127.0.0.1 server 0.uk.pool.ntp.org server 1.uk.pool.ntp.org server 2.uk.pool.ntp.org server 3.uk.pool.ntp.org driftfile /etc/ntp/drift broadcastdelay 0.008 authenticate yes keys/etc/ntp/keys # /etc/init.d/ntpd start Starting ntpd: [ OK ] # grep ntp /var/log/messages Sep 4 15:00:40 amber ntpd: ntpd startup succeeded Sep 4 15:00:40 amber ntpd[3083]: ntpd 4.1...@1.791 Sat Aug 31 18:27:29 EDT 2002 (1) Sep 4 15:00:40 amber ntpd[3083]: precision = 13 usec Sep 4 15:00:40 amber ntpd[3083]: kernel time discipline status 0040 Sep 4 15:00:40 amber ntpd[3083]: frequency initialized 0.000 from /etc/ntp/drift # ntpq -p remote refid st t when poll reach delay offset jitter == ntp0.cis.strath 0.0.0.0 16 u- 6400.0000.000 4000.00 ntp1.exa.net.uk 0.0.0.0 16 u- 6400.0000.000 4000.00 dns1.rmplc.co.u 0.0.0.0 16 u- 6400.0000.000 4000.00 odie.tomelliott 0.0.0.0 16 u- 6400.0000.000 4000.00 # date Fri Sep 4 15:40:21 BST 2009 # ntpq -p remote refid st t when poll reach delay offset jitter == ntp0.cis.strath 0.0.0.0 16 u- 12800.0000.000 4000.00 ntp1.exa.net.uk 0.0.0.0 16 u- 12800.0000.000 4000.00 dns1.rmplc.co.u 0.0.0.0 16 u- 12800.0000.000 4000.00 odie.tomelliott 0.0.0.0 16 u- 12800.0000.000 4000.00 # ntpdate -q ntp1.exa.net.uk server 82.219.4.30, stratum 2, offset -0.061504, delay 0.07434 4 Sep 15:40:57 ntpdate[3311]: adjust time server 82.219.4.30 offset -0.061504 sec So, no contact with NTP servers after 40 mins of running, even though ntpdate can query those same servers quite happily. Any ideas how I can find out what's wrong? Lose the restrict statements in ntp.conf. You should then be able to receive replies from the servers! Restrict is not as well documented as it might be! You told ntpd to ignore ALL servers with restrict default ignore ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
[ntp:questions] ntp server and client
Hi All I just need to ask a quick and maybe a dumb one, sorry but I do not have time to search for the solution on the net. I want to set up a NetBSD unit to run as ntp server and client. At the moment my configuration file looks like this: ### server 127.127.1.0 fudge 127.127.1.0 stratum 4 server 0.europe.pool.ntp.org iburst server 1.europe.pool.ntp.org iburst server 0.us.pool.ntp.org iburst server 1.us.pool.ntp.org iburst ### A the moment my local unit is syncing to the local clock? How do I change this so that I give time updates to local clients, and get the time from the external stratum 2 server? Thanks Jason ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
[ntp:questions] ntpd reads from shared memory but doesn't act on it
Hi, I'm trying to get ntpd to read from shared memory and then adjust the time accordingly. Problem is it seems to read the value that I put into shared memory but doesn't seem to adjust the time accordingly. This is how I ran ntpd: ./ntpd -D 3 -u 113:124 -g Using the gpsd program, I borrowed code from there and wrote my own little program to write into shared memory. When I run my program, I have it set to increase the time by 60 seconds and 300 microseconds, i.e., 60.000300 seconds. Example output from my program: ntpshm_put: Clock: 1251851559 @ 1251851499.390156 Additionally, there is an entry in my /etc/ntp.conf that says fudge is 0.420. When I run my program, I see this output from gpsd immediately: peer SHM(0) event 'event_reach' (0x84) status 'unreach, conf, 2 events, event_reach' (0x8024) refclock_sample: n 1 offset 60.420300 disp 0.00 jitter 0.01 clock_filter: n 1 off 60.420300 del 0.00 dsp 7.937500 jit 0.01, age 0 That offset - 60.420300 - seems consistent with what I set. (60.0003000 + 0.420) However even after half a day, ntpd doesn't seem to have adjusted the time. Maybe I'm forgetting to set some option? Maybe once is not enough for ntpd, and I need to keep writing new values regularly? Any idea on what I might be forgetting or doing incorrectly? Some relevant information: I don't want my ntpd to talk to external servers. The intention is that a script will write values into shared memory and ntpd should read this and adjust accordingly. This is what my /etc/ntp.conf file looks like: driftfile /var/lib/ntp/ntp.drift # Enable this if you want statistics to be logged. statsdir /var/log/ntpstats/ statistics loopstats peerstats clockstats filegen loopstats file loopstats type day enable filegen peerstats file peerstats type day enable filegen clockstats file clockstats type day enable # You do need to talk to an NTP server or two (or three). #server ntp.ubuntu.com # By default, exchange time with everybody, but don't allow configuration. restrict -4 default kod notrap nomodify nopeer noquery restrict -6 default kod notrap nomodify nopeer noquery # Local users may interrogate the ntp server more closely. restrict 127.0.0.1 restrict ::1 # Clients from this (example!) subnet have unlimited access, but only if # cryptographically authenticated. #restrict 192.168.123.0 mask 255.255.255.0 notrust server 127.127.28.0 mode 1 minpoll 4 maxpoll 4 fudge 127.127.28.0 time1 0.420 refid GPS server 127.127.28.1 mode 1 minpoll 4 maxpoll 4 prefer fudge 127.127.28.1 refid GPS1 The output of ntpq -p remote refid st t when poll reach delay offset jitter == SHM(0) .GPS.0 l 17h 1600.0000.000 0.001 SHM(1) .GPS1. 0 l- 1600.0000.000 0.001 That 17h corresponds to when my program wrote into the shared memory. The output of ntpdc ntpdc loopinfo offset: 0.00 s frequency:-34.340 ppm poll adjust: 0 watchdog timer: 63469 s $ cat /var/lib/ntp/ntp.drift -34.340 $ cat /proc/sysvipc/shm key shmid perms size cpid lpid nattch uid gid cuid cgid atime dtime ctime 1314148400 0700 80 5014 25550 20 00 0 1251851499 1251851500 1251847089 131414840132769 70080 5014 25550 100 0 0 1251851499 1251851500 1251847089 0 65538 1600 393216 5466 55882 106 114 106 114 1251847098 1251847098 1251847098 On strange thing I did notice is that the lpid in the above output is always that of my program which writes into it. Looking at the ntpd output, it seems to read it from shared memory *after* my program has written into it. So, shouldn't the lpid be that of the ntpd process? FWIW, I'm using ntpd version 4.2.4p7 on a 32-bit Ubuntu 8.04 machine. Thanks. _ With Windows Live, you can organize, edit, and share your photos. http://www.windowslive.com/Desktop/PhotoGallery ___ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
