13 Oct
2022
13 Oct
'22
3:28 a.m.
Hey KI5PGJ,
Interesting to see you only got "some" Mbps. We have tested this setup many
times, and I'm pretty sure the bottleneck is around 2.5~3.5 Gbps via the
browser.
These numbers defeat the purpose and are a lot higher than your results.
Indeed, I would rather suspect the proxy, or something else between your
test machine and the 44.137.0.0/16 infrastructure :)
I'm somewhere curious to know which route your connection is running
through! Do you happen to have a traceroute at your fingertips :)?
73, PH4T
On Wed, 12 Oct 2022 at 17:42, Rob PE1CHL via 44net <44net(a)mailman.ampr.org>
wrote:
Ok... thanks for trying. Our gateway is hosted at
SURF, the educational
network provider here in the Netherlands.
A user within that same network got about 3 Gbps on the speedtest, so at
least I know it is not limited to 1 Gbps somehow (it used to be due to a 1
Gbps link between the servers, but that has since been reconfigured).
It could be we are limited by CPU speed of either the speedtest server
itself or the MikroTik CHR that is in front of it (which has a 10 Gbps
license).
Locally on another VM I do see speed up to about 5.5 Gbps, but that was
tested using iperf3. The browser and any proxy can also play a role, yes.
Anyway.... fast enough for AMPRNet :-) We typically have 40-80 Mbps
traffic only.
Rob
On 10/12/22 16:08, KI5PGJ via 44net wrote:
_______________________________________________
44net mailing list -- 44net(a)mailman.ampr.org
To unsubscribe send an email to 44net-leave(a)mailman.ampr.org
I ran the speed test from work via our 10G ESNet
connection, below is
the results link. The latency is pretty high because of the
distance, also
our proxy most likely had an impact too.