All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Hezner upload speeds in Finland
Anyone else having bad upload speed in Finland with Hezner?
root@ # speedtest-cli
Retrieving speedtest.net configuration...
Testing from Hetzner Online GmbH ...
Retrieving speedtest.net server list...
Selecting best server based on ping...
Hosted by Nebula Oy (Helsinki) [0.18 km]: 8.535 ms
Testing download
speed................................................................................
Download: 874.55 Mbit/s
Testing upload
speed......................................................................................................
Upload: 4.15 Mbit/s
root@# speedtest-cli
Retrieving speedtest.net configuration...
Testing from Hetzner Online GmbH ...
Retrieving speedtest.net server list...
Selecting best server based on ping...
Hosted by Cumulus Technologies (Helsinki) [0.18 km]: 4.381 ms
Testing download
speed................................................................................
Download: 767.98 Mbit/s
Testing upload
speed......................................................................................................
Upload: 4.15 Mbit/s
root@ # speedtest-cli
Retrieving speedtest.net configuration...
Testing from Hetzner Online GmbH ...
Retrieving speedtest.net server list...
Selecting best server based on ping...
Hosted by Cumulus Technologies (Helsinki) [0.18 km]: 4.326 ms
Testing download
speed................................................................................
Download: 609.32 Mbit/s
Testing upload
speed......................................................................................................
Upload: 4.15 Mbit/s
root@# speedtest-cli --share
Retrieving speedtest.net configuration...
Testing from Hetzner Online GmbH ...
Retrieving speedtest.net server list...
Selecting best server based on ping...
Hosted by Cumulus Technologies (Helsinki) [0.18 km]: 4.539 ms
Testing download
speed................................................................................
Download: 775.76 Mbit/s
Testing upload
speed......................................................................................................
Upload: 4.15 Mbit/s
Share results:
root@~ # speedtest-cli --share
Retrieving speedtest.net configuration...
Testing from Hetzner Online GmbH ..
Retrieving speedtest.net server list...
Selecting best server based on ping...
Hosted by Cumulus Technologies (Helsinki) [0.18 km]: 4.282 ms
Testing download
speed................................................................................
Download: 740.30 Mbit/s
Testing upload
speed......................................................................................................
Upload: 4.15 Mbit/s
Share results:
root@~ # speedtest-cli --share
Retrieving speedtest.net configuration...
Testing from Hetzner Online GmbH ...
Retrieving speedtest.net server list...
Selecting best server based on ping...
Hosted by Cumulus Technologies (Helsinki) [0.18 km]: 4.59 ms
Testing download
speed................................................................................
Download: 769.15 Mbit/s
Testing upload
speed......................................................................................................
Upload: 4.15 Mbit/s
Share results:
root@~ #
Comments
That's probably the same
speedtest-cli
bug as mentioned here: https://www.lowendtalk.com/discussion/comment/3065318/#Comment_3065318Disable BBR and use offical app.
Try https://www.speedtest.net/apps/cli
Now the speeds are showing correct thanks !
Wow significant difference in speedtest-cli versus official speedtest in cli mode
post a YABS please? to see finland speeds
yabs ?
Yep, the dev of speedtest-cli on github didn't want to fix it. So speedtest built their own. Which now works.
https://github.com/masonr/yet-another-bench-script
This is the iperf section of YABS on one of my Finland EX42's. Note this is a fully configured server and not a fresh box, so other things may have been using the network during the test.
Check if you have full duplex on the network card; maybe a bad cable?
I get the same sort of figures on a second EX42 also in Finland. Both report 1000Mb/s full duplex...
For the record, the machines are Debian Buster with Proxmox installed and running a few VM's each.
Basic System Information:
Processor : Intel(R) Xeon(R) CPU E3-1275 v6 @ 3.80GHz
CPU cores : 8 @ 1364.963 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 62Gi
Swap : 31Gi
Disk : 437G
dd Sequential Disk Speed Tests:
Write | 568 MB/s | 595 MB/s | 594 MB/s | 585.67 MB/s
Read | 2.6 GB/s | 2.7 GB/s | 2.9 GB/s | 2.73 GB/s
iperf3 Network Speed Tests (IPv4):
Provider | Location (Link) | Send Speed | Recv Speed
| | |
Bouygues Telecom | Paris, FR (10G) | 475 Mbits/sec | 917 Mbits/sec
Online.net | Paris, FR (10G) | 476 Mbits/sec | 929 Mbits/sec
Severius | The Netherlands (10G) | busy bits/sec | busy
Worldstream | The Netherlands (10G) | 477 Mbits/sec | 931 Mbits/sec
wilhelm.tel | Hamburg, DE (10G) | 484 Mbits/sec | 931 Mbits/sec
Biznet | Bogor, Indonesia (1G) | 388 Mbits/sec | 189 Mbits/sec
Hostkey | Moscow, RU (1G) | 490 Mbits/sec | 938 Mbits/sec
Vultr | Piscataway, NJ, US (1G) | busy bits/sec | busy bits/sec
Velocity Online | Tallahassee, FL, US (10G) | 432 Mbits/sec | 452 Mbits/sec
Airstream Communications | Eau Claire, WI, US (10G) | 417 Mbits/sec | 348 Mbits/sec
Hurricane Electric | Fremont, CA, US (10G) | busy | 186 Mbits/sec
iperf3 Network Speed Tests (IPv6):
Provider | Location (Link) | Send Speed | Recv Speed
| | |
Bouygues Telecom | Paris, FR (10G) | 470 Mbits/sec | 911 Mbits/sec
Online.net | Paris, FR (10G) | 475 Mbits/sec | 912 Mbits/sec
Severius | The Netherlands (10G) | busy | busy bits/sec
Worldstream | The Netherlands (10G) | 479 Mbits/sec | 911 Mbits/sec
wilhelm.tel | Hamburg, DE (10G) | 494 Mbits/sec | 911 Mbits/sec
Vultr | Piscataway, NJ, US (1G) | busy bits/sec | busy bits/sec
Airstream Communications | Eau Claire, WI, US (10G) | 414 Mbits/sec | 499 Mbits/sec
Hurricane Electric | Fremont, CA, US (10G) | 426 Mbits/sec | 319 Mbits/sec
Geekbench 4 Benchmark Test:
Test | Value
|
Single Core | 5298
Multi Core | 18479
Full Test | https://browser.geekbench.com/v4/cpu/15189083
Got it; just a wild guess, pity it’s not the s tusk cause as it would be easily fixable.
I guess the only option is to leave it with their support.
No such thing as half duplex or a bad duplex setting on a gigabit nic. 1000 base-t is auto nego only.
Not sure what NIC you have but this helped on my Hetzner server with the Intel I219-LM along with BBR.
https://wiki.hetzner.de/index.php/Low_performance_with_Intel_i218/i219_NIC/en
The ethtool thing certainly improved things
Now does that stick over a reboot, or do I need to do it at boot?
Also, BBR?
At each boot - unless someone knows how to keep it permanent?
Enable TCP BBR : Just give it quick google