Slow performance with bridges

[KevinAls]

My router can route directly via upstream ISP or via GRE tunnel. So in this case, proxmox has nothing to do with it.

As long as I dont route my traffic via the GRE, all my VM's can do full power on bandwidth. Whenever I switch to the GRE, debian boxes not running proxmox goes full speed, but everything running proxmox (and thus, also all vm's under proxmox) is running with 50-60mbits.

What could this be caused? I dont think MTU is wrong as the barebones servers running Ubuntu/Debian runs just fine.

Thank you!

 

[KevinAls]

To give some more info:
Barebones iperf to hetzner:
[ ID] Interval Transfer Bitrate Retr Cwnd [ 5] 0.00-1.00 sec 24.0 MBytes 201 Mbits/sec 154 549 KBytes [ 5] 1.00-2.00 sec 23.8 MBytes 199 Mbits/sec 12 564 KBytes [ 5] 2.00-3.00 sec 23.8 MBytes 199 Mbits/sec 43 553 KBytes [ 5] 3.00-4.00 sec 23.8 MBytes 199 Mbits/sec 89 559 KBytes [ 5] 4.00-5.00 sec 23.8 MBytes 199 Mbits/sec 55 544 KBytes [ 5] 5.00-6.00 sec 22.5 MBytes 189 Mbits/sec 6 522 KBytes [ 5] 6.00-7.00 sec 27.5 MBytes 231 Mbits/sec 31 658 KBytes [ 5] 7.00-8.00 sec 28.8 MBytes 241 Mbits/sec 27 663 KBytes [ 5] 8.00-9.00 sec 23.8 MBytes 199 Mbits/sec 44 663 KBytes [ 5] 9.00-10.00 sec 27.5 MBytes 231 Mbits/sec 22 655 KBytes - - - - - - - - - - - - - - - - - - - - - - - - - [ ID] Interval Transfer Bitrate Retr [ 5] 0.00-10.00 sec 249 MBytes 209 Mbits/sec 483 sender [ 5] 0.00-10.02 sec 247 MBytes 207 Mbits/sec receiver

Proxmox (host) to same hetzner:
[ ID] Interval Transfer Bitrate Retr Cwnd [ 5] 0.00-1.00 sec 7.43 MBytes 62.3 Mbits/sec 48 126 KBytes [ 5] 1.00-2.00 sec 7.39 MBytes 62.0 Mbits/sec 0 165 KBytes [ 5] 2.00-3.00 sec 7.89 MBytes 66.2 Mbits/sec 4 102 KBytes [ 5] 3.00-4.00 sec 4.35 MBytes 36.5 Mbits/sec 13 73.5 KBytes [ 5] 4.00-5.00 sec 3.98 MBytes 33.4 Mbits/sec 1 80.6 KBytes [ 5] 5.00-6.00 sec 4.35 MBytes 36.5 Mbits/sec 5 86.3 KBytes [ 5] 6.00-7.00 sec 3.91 MBytes 32.8 Mbits/sec 2 66.5 KBytes [ 5] 7.00-8.00 sec 4.35 MBytes 36.5 Mbits/sec 0 105 KBytes [ 5] 8.00-9.00 sec 6.52 MBytes 54.7 Mbits/sec 0 144 KBytes [ 5] 9.00-10.00 sec 5.65 MBytes 47.4 Mbits/sec 4 97.6 KBytes - - - - - - - - - - - - - - - - - - - - - - - - - [ ID] Interval Transfer Bitrate Retr [ 5] 0.00-10.00 sec 55.8 MBytes 46.8 Mbits/sec 77 sender [ 5] 0.00-10.03 sec 55.1 MBytes 46.0 Mbits/sec receiver

Both servers are on the same switch. Its both DL360 G9 servers

 

[boxofrocks]

You have a Congestion window problem. Meaning something is slowing you down between that machine and the endpoints. CWR could be ICMP not being allowed over the GRE. This would then cause a problem if there was a network slow down and the two end points had to hold/buffer data. They don't have a fast re-open if the MTU and link speed is small. Need to see what your consistent packet size is vs the average or working packets.

 

[KevinAls]

You have a Congestion window problem. Meaning something is slowing you down between that machine and the endpoints. CWR could be ICMP not being allowed over the GRE. This would then cause a problem if there was a network slow down and the two end points had to hold/buffer data. They don't have a fast re-open if the MTU and link speed is small. Need to see what your consistent packet size is vs the average or working packets.

Thank you very much for your answer.

I've been playing around with different MTU sizes, however, the issue still persists.

It doesnt really answer why the barebones one always just works.

 

[KevinAls]

You have a Congestion window problem. Meaning something is slowing you down between that machine and the endpoints. CWR could be ICMP not being allowed over the GRE. This would then cause a problem if there was a network slow down and the two end points had to hold/buffer data. They don't have a fast re-open if the MTU and link speed is small. Need to see what your consistent packet size is vs the average or working packets.

Link speed: 500/500
MTU: 1476 (1452 also tried)
TCP clamping is enabled.

ICMP is also allowed via the GRE.