3 ms·

Out of curiosity, have you analyzed how different/better your setup is compared to just using something like time.google.com?

magicalhippo 5y ago
I admit it was more for the geek factor than anything else, and I'm no expert in the field. Here are the source stats[1] from chrony: Pi with GPS $ chronyc sourcestats 210 Number of sources = 7 Name/IP Address NP NR Span Frequency Freq Skew Offset Std Dev ============================================================================== PPS1 8 6 112 -0.018 0.020 -105ns 336ns GPS1 7 6 95 +55.685 111.018 +126ms 1058us GPS2 8 6 112 -0.018 0.021 -89ns 341ns 2a01:79a:408::250 18 10 310m -0.057 0.080 -1364us 522us time.cloudflare.com 19 13 344m -0.068 0.038 -1274us 258us ntp.coreless.net 6 3 85m -0.343 0.706 -2355us 354us anduin.net 21 8 379m -0.020 0.051 -835us 388us Debian VM running on Bhyve(FreeBSD) in same LAN $ chronyc sourcestats 210 Number of sources = 5 Name/IP Address NP NR Span Frequency Freq Skew Offset Std Dev ============================================================================== timepi.localdomain 13 7 64m +0.001 0.031 +328ns 28us 79.161.39.253.static.lys> 8 6 120m -0.002 0.184 -2122us 189us 125.51-174-177.customer.> 22 13 464m +0.287 0.117 -839us 1201us time.cloudflare.com 6 5 85m +0.009 0.472 -1615us 277us c1EE4653E.static.as2116.> 10 7 155m -0.146 0.603 -3315us 1256us So on the VM it seems to "just" be an order of magnitude better. But that should be worst-case. [1]: https://chrony.tuxfamily.org/doc/3.3/chronyc.html https://chrony.tuxfamily.org/doc/3.3/chronyc.html
- jorangreef 5y ago
  Nice!

magicalhippo 5y ago

I admit it was more for the geek factor than anything else, and I'm no expert in the field. Here are the source stats[1] from chrony: Pi with GPS $ chronyc sourcestats 210 Number of sources = 7 Name/IP Address NP NR Span Frequency Freq Skew Offset Std Dev ============================================================================== PPS1 8 6 112 -0.018 0.020 -105ns 336ns GPS1 7 6 95 +55.685 111.018 +126ms 1058us GPS2 8 6 112 -0.018 0.021 -89ns 341ns 2a01:79a:408::250 18 10 310m -0.057 0.080 -1364us 522us time.cloudflare.com 19 13 344m -0.068 0.038 -1274us 258us ntp.coreless.net 6 3 85m -0.343 0.706 -2355us 354us anduin.net 21 8 379m -0.020 0.051 -835us 388us Debian VM running on Bhyve(FreeBSD) in same LAN $ chronyc sourcestats 210 Number of sources = 5 Name/IP Address NP NR Span Frequency Freq Skew Offset Std Dev ============================================================================== timepi.localdomain 13 7 64m +0.001 0.031 +328ns 28us 79.161.39.253.static.lys> 8 6 120m -0.002 0.184 -2122us 189us 125.51-174-177.customer.> 22 13 464m +0.287 0.117 -839us 1201us time.cloudflare.com 6 5 85m +0.009 0.472 -1615us 277us c1EE4653E.static.as2116.> 10 7 155m -0.146 0.603 -3315us 1256us So on the VM it seems to "just" be an order of magnitude better. But that should be worst-case. [1]: https://chrony.tuxfamily.org/doc/3.3/chronyc.html https://chrony.tuxfamily.org/doc/3.3/chronyc.html

jorangreef 5y ago
Nice!