On 08.06.2018 12:21, Ralf Hildebrandt wrote:
Beware, UDP is limited to 1 Mbit in iperf ("default 1 Mbit/sec for UDP, unlimited for TCP"), thus:iperf3 --udp --bandwidth 200M -c 172.31.254.1 Also, check CPU on both machines to see if you're CPU bound.
I found what with argument "--parallel" I see normal throughput: # iperf3 -c 172.31.254.1 --parallel 8 [SUM] 0.00-10.00 sec 107 MBytes 89.4 Mbits/sec sender [SUM] 0.00-10.00 sec 105 MBytes 87.7 Mbits/sec receiver # iperf3 -c 172.31.254.1 --parallel 8 --reverse [SUM] 0.00-10.00 sec 108 MBytes 90.5 Mbits/sec sender [SUM] 0.00-10.00 sec 105 MBytes 88.3 Mbits/sec receiver # iperf3 -c 172.31.254.1 --parallel 16 [SUM] 0.00-10.00 sec 108 MBytes 90.8 Mbits/sec sender [SUM] 0.00-10.00 sec 105 MBytes 88.5 Mbits/sec receiver # iperf3 -c 172.31.254.1 --parallel 16 --reverse [SUM] 0.00-10.00 sec 108 MBytes 90.5 Mbits/sec sender [SUM] 0.00-10.00 sec 105 MBytes 88.1 Mbits/sec receiver In real-world usage it will be many different TCP connections, so all is Ok with OpenVPN - VPS bandwidth is limited to 100 Mbit/sec. First idea: probably hosting or network providers just limit maximum allowed throughput for one TCP connection? But wait, this is TCP connections *inside* OpenVPN tunnel. So, on network layer it will be just flow of UDP packets. I can't understand the root cause of observed behavio(u)r. I can't find any explanation why one TCP connection is so slow. May be something wrong with settings of operating systems? It is CentOS 7.5 64 bit with all default sysctl settings, I just change only one setting: "net.ipv4.ip_forward = 1" -- Best regards, Gena ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Openvpn-users mailing list Openvpn-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/openvpn-users
