I wish I had those results of yours!I think there's a growing reliance on perceived performance based on latency that doesn't mean anything. Of course there are going to be differences based on location. That's inherently how it works.
The chances of a data center that hosts Cloudflare or Google DNS near Somerset are slim, and your latency reflects that. Where as I, being in Kent, am significantly closer to those larger sites, whether Slough or Docklands, which is more than likely where these are. Manchester as well.
That said, latency to a DNS provider is a very small piece of the puzzle.
As proof to that end, I'm 2.3 and 1.6ms RTT to the above providers respectively.
On the contrary, I'd argue that latency tells much more of a story than you are give it credit for. Firstly, if you are constantly testing against the same location, then any variability in the latencies would often highlight bad traffic management, overwhelmed pipes or routing anomalies.I think there's a growing reliance on perceived performance based on latency that doesn't mean anything. Of course there are going to be differences based on location. That's inherently how it works.
That's just it...you have to get to those locations. As mentioned, Somerset is nowhere near, so comparing my results to the OP's results are irrelevant. We're no where near each other geographically so you can't possibly compare the two. Then, you need to know who/where is actually responding. Yes, Cloudflare has presence in those locations. But you have no way to guarantee who is answering at any given point in time. All you know is that ICMP, which is terrible at validating anything in any consistent way, had some result at that point in time. What HW is the user using? What load on their own network exists at the time? Any variability with that throws results out the window. There is no control over how anyone tests, what HW they use, where they are, who is responding, or any other issue between point A and B that may or may not exist at the time.On the contrary, I'd argue that latency tells much more of a story than you are give it credit for. Firstly, if you are constantly testing against the same location, then any variability in the latencies would often highlight bad traffic management, overwhelmed pipes or routing anomalies.
Secondly, in the case of YouFibre, testing against Cloudflare is a pretty good choice for roughly measuring latency to YouFibre's main transit boundaries. The bulk of YouFibre's internet routing goes via two sites in London and one in Manchester, and Cloudflare have direct presence/peerings with YouFibre and anycast their DNS service in those locations, so you can get a pretty good idea of what your baseline latency is to cross the YouFibre network itself.
https://www.thinkbroadband.com/broa...hare/ab91eca9e21e26fb477a617d85df33a1f581fc90
On the 8000mbps service.
Nice. What software is that? ThanksHere is mine, I'm on the 1GB service.
It's the Unifi console, I have one of their devices.Nice. What software is that? Thanks
Reminded me, also have the same.I don't have a static IPv4 and run tbb monitor to my IPv6, when I first signed up the graph was flat smooth but then it went weird sometime last summer and it gets a couple of spikes to 20ms in every minute (confirmed by pinging various IPv6 enabled sites myself) so the graph looks like this, not too sure why.
That latency is insane. You must be plugged into the server!Reminded me, also have the same.
It's impressive eh, wonder if take up is low here and that is why? It was completed only a few months ago.That latency is insane. You must be plugged into the server!... I wish
Just wondered if other YF users on static ipv4 care to share latency charts and approx location.
I'll start
2gb/2gb - somerset currently latency around 7.5ms (down from recent longterm figure of 11ms)
ipv4
ipv6
Internet speed in general attachedView attachment 11975
thanks
Hi. I'm using a debian box with docker and portainer. A speedtest app sits in that which outputs to influx. Grafana is using a custom dashboard using that data.Rather know peoples waveform tests lol
what graphana tests you doing o.0
Don't really see any. WAN never saturates other than under artificial conditions and not bothered about thatRather know peoples waveform tests lol