NTP widget time wrong?
-
@Ramosel said in NTP widget time wrong?:
time-c-b.nist.gov
Just tried adding that...somehow pfSense is not seeing it as a valid address after clicking save. I take it the "c" is for central time.
Tried the IP and it took that...
pfSense+ 23.09 Lenovo Thinkcentre M93P SFF Quadcore i7 dual Raid-ZFS 128GB-SSD 32GB-RAM PCI-Intel i350-t4 NIC, -Intel QAT 8950.
pfSense+ 23.09 VM-Proxmox, Dell Precision Xeon-W2155 Nvme 500GB-ZFS 128GB-RAM PCIe-Intel i350-t4, Intel QAT-8950, P-cloud. -
@NollipfSense time-c-g.nist.gov resolves just fine. But you shouldn't be marking things as pools when they are not pools.
;; QUESTION SECTION: ;time-c-g.nist.gov. IN A ;; ANSWER SECTION: time-c-g.nist.gov. 3600 IN A 129.6.15.30https://tf.nist.gov/tf-cgi/servers.cgi
Looks like you have a space on the end there as well..
-
@johnpoz Thanks John, I'll fix that.
That worked...
-
@NollipfSense said in NTP widget time wrong?:
@Ramosel said in NTP widget time wrong?:
time-c-b.nist.gov
Just tried adding that...somehow pfSense is not seeing it as a valid address after clicking save. I take it the "c" is for central time.
No, it's just a lesser used IPV4 server in Boulder. Dave MIlls is an old acquaintance and when I spoke with him about my pfSense setup a few years back he said the a-b servers get hammered the most... all the IPV6 guys flock to the d-b and e-b servers... use c-b.
https://tf.nist.gov/tf-cgi/servers.cgi
-
@Ramosel
That is actually not a bad price for the "Hat" ... €50 from France.Right now i'm doing a DIY NTP server, with an OrangePI One.
That will be built in my - RAPCO 1804 L25C GPSDO Receiver
I have chosen the OrangePi over the RasPi's (I have 10+ 3B's in the drawer), due to the Lan CHIP being connected directly to the SoC.
The RasPi 3B Lan chip is actually USB coupled, and that causes a bit of jitter.This is my current "burn in" test of the OrangePi.
With a DS3231 IIC Module - https://www.aliexpress.com/item/32822420722.html
$ sudo i2cdetect -y 1 0 1 2 3 4 5 6 7 8 9 a b c d e f 00: -- -- -- -- -- -- -- -- 10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 30: 30 -- 32 33 34 35 36 -- -- -- -- -- -- -- -- -- 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 50: 50 51 52 53 54 55 56 57 58 59 5a 5b 5c 5d 5e 5f 60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 70: -- -- -- -- -- -- -- --$ sudo hwclock --verbose hwclock from util-linux 2.36.1 System Time: 1695225080.007973 Trying to open: /dev/rtc0 Using the rtc interface to the clock. Last drift adjustment done at 1695223549 seconds after 1969 Last calibration done at 1695223549 seconds after 1969 Hardware clock is on UTC time Assuming hardware clock is kept in UTC time. Waiting for clock tick... ...got clock tick Time read from Hardware Clock: 2023/09/20 15:51:41 Hw clock time : 2023/09/20 15:51:41 = 1695225101 seconds since 1969 Time since last adjustment is 1552 seconds Calculated Hardware Clock drift is 0.000000 seconds 2023-09-20 17:51:39.554254+02:00The hardware will be "transplanted" into the GPSDO Box , and i will piggyback on the existing Trimble GPS receiver (RX + 1PPS).
Then install & run GPSD + NTP w. 1-PPS support/Bingo
-
@bingo600 yes, that is what I mean to NAT to your NTP server's IP address for clients not manually going to it. For me I use authenticated NTP so the firewall points to those when requests hit the firewall.
-
@bingo600 said in NTP widget time wrong?:
@Ramosel
That is actually not a bad price for the "Hat" ... €50 from France.Right now i'm doing a DIY NTP server, with an OrangePI One.
That will be built in my - RAPCO 1804 L25C GPSDO ReceiverI like that, sweet. Thanks for the info. Most of these builds are turning off the USB and using GPIO. The i2c gets really simple. Phil gets really deep into fine tuning after you get it running... I need to go back and do more, but I though the Pi NTP server was causing the widget issue. Now that I know it's not, I'll dig deeper into the GPS/NTP/PPS dialog.
pi@RPi4b-NTP-GPS:~ $ i2cdetect -y 1
0 1 2 3 4 5 6 7 8 9 a b c d e f
00: -- -- -- -- -- -- -- --
10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
40: -- -- 42 -- -- -- -- -- -- -- -- -- -- -- -- --
50: -- -- UU -- -- -- -- -- -- -- -- -- -- -- -- --
60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
70: -- -- -- -- -- -- -- --I'm curious if Phil's build would work on the Orange Pi??
http://www.philrandal.co.uk/blog/archives/2019/04/entry_213.html -
Cannot get the devices to agree on time, one will be -0.046s while another device will be +0.045s
-
@Ramosel said in NTP widget time wrong?:
I'm curious if Phil's build would work on the Orange Pi??
http://www.philrandal.co.uk/blog/archives/2019/04/entry_213.htmlThe Trimble TSIP protocol is somewhat special, compared to NMEA.
But i'm in contact with the maintainer of the TSIP part of GPSD, and he's quite sure GPSD will work.
Even with a receive only (Uart RX + 1-PPS) connection. I can't (won't) connect TX to the OrangePI, as that "belongs to the Rapco".If i go totally crazy i could test for "missing 1-PPS" for ie. 20 sec , and then switch over to a 1-PPS divided down from the 10MHz OCXO in the Rapco.
That way "I'll be somewhat in sync .. Depends on the holdover routine in the Rapco" even with a GPS outage.Hmmm .... Maybe the HW dependent part should be moved to the "Off Topic thread" you made.
https://forum.netgate.com/topic/182811/raspberry-gps-based-time-server/Bingo
-
@bingo600 said in NTP widget time wrong?:
Hmmm .... Maybe the HW dependent part should be moved to the "Off Topic thread" you made.
https://forum.netgate.com/topic/182811/raspberry-gps-based-time-serverSure, now that I know I'm not crazy and the NTP clock is really grunged... I'm good here. I doubt that I'd ever build what you have, but you've got me interested and I'd sure like to follow the build. Use my thread or start your own, I'm cool either way.
Rick
