Hello all,
My apologies if this question has already been answered but I am having troubles finding the answer.
I am currently testing SoftEther in a VM to VM environment to see the maximum performance I would be able to get out of it. When testing with a single core server I can put about 500mbps through a VPN with UDP packets.
I then increased to 2 CPU's and rebooted the server. During the tests no performance was gained and the second CPU was not utilised. This leads to the following question,is SoftEther capable of multithreading?
If so, does the second CPU core has to be present when building or is that irrelevant.
Any advice on how performance can be increased is also welcome, I am currently using a local bridge with NAT as setup.
Many thanks in advance,
Regards,
/dev/null
EDIT:
1. Operating system name and the type of CPU-bits
Debian 8.2 Jessie x64
2. The result of "ifconfig –a" (UNIX) or "ipconfig /all" (Windows)
eth0 Link encap:Ethernet HWaddr 08:00:27:d7:44:9c
inet addr:192.168.56.103 Bcast:192.168.56.255 Mask:255.255.255.0
inet6 addr: fe80::a00:27ff:fed7:449c/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:15704726 errors:0 dropped:0 overruns:0 frame:0
TX packets:5692 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:12959891433 (12.0 GiB) TX bytes:1135945 (1.0 MiB)
eth1 Link encap:Ethernet HWaddr 08:00:27:25:df:ef
inet addr:192.168.2.46 Bcast:192.168.2.255 Mask:255.255.255.0
inet6 addr: fe80::a00:27ff:fe25:dfef/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:3550 errors:0 dropped:0 overruns:0 frame:0
TX packets:1645 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:274715 (268.2 KiB) TX bytes:382594 (373.6 KiB)
lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
inet6 addr: ::1/128 Scope:Host
UP LOOPBACK RUNNING MTU:65536 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:0 (0.0 B) TX bytes:0 (0.0 B)
tap_dhcp Link encap:Ethernet HWaddr 00:ac:d1:36:a1:42
inet addr:192.168.7.1 Bcast:192.168.7.255 Mask:255.255.255.0
inet6 addr: fe80::2ac:d1ff:fe36:a142/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:7789137 errors:0 dropped:0 overruns:0 frame:0
TX packets:1538 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:500
RX bytes:11775642432 (10.9 GiB) TX bytes:333774 (325.9 KiB)
3. The result of "uname –a" (UNIX) or "systeminfo" (Windows)
Linux debian 3.16.0-4-amd64 #1 SMP Debian 3.16.7-ckt11-1+deb8u5 (2015-10-09) x86_64 GNU/Linux
4. The build number of SoftEther VPN
Latest version on the website
5. Which SoftEther VPN component are you using?
Client for Windows and server for Linux
6. Whether or not there is a NAT or Firewall between your VPN server and the Internet.
VM to VM so nothing in between
7. Are you using SecureNAT?
NO because of performance hits
8. Your current vpn_server.config or vpn_bridge.config file should be attached on the post.
This shall be posted if necessary
Multithread
-
- Posts: 2458
- Joined: Mon Feb 24, 2014 11:03 am
Re: Multithread
Please try TCP connection and increase TCP connections.
-
- Site Admin
- Posts: 2081
- Joined: Sat Mar 09, 2013 5:37 am
Re: Multithread
Threads are assigned for each session.
If there is only single session, the VPN server can't use multiple cores efficiently.
If there is only single session, the VPN server can't use multiple cores efficiently.
-
- Posts: 6
- Joined: Thu Sep 15, 2016 2:22 pm
Re: Multithread
Firstly, thank you to everyone involved with this project for an excellent piece of software!
I'm sorry to resurrect an old thread, however I noticed that with compression enabled the CPU use seems to max out at 130% on a quad-core device running vpnserver, i.e. it looks that it doesn't scale well across the cores.
Test conditions - both ends running on Linux, performance testing with iperf and two clients and two servers at each end, both clients running "-w 256k -d". Cascade connection configured for 16 sessions. Throughput is the same with one or two clients running, about 55Mbps in total. Inter-site latency is about 15ms.
I'm sorry to resurrect an old thread, however I noticed that with compression enabled the CPU use seems to max out at 130% on a quad-core device running vpnserver, i.e. it looks that it doesn't scale well across the cores.
Test conditions - both ends running on Linux, performance testing with iperf and two clients and two servers at each end, both clients running "-w 256k -d". Cascade connection configured for 16 sessions. Throughput is the same with one or two clients running, about 55Mbps in total. Inter-site latency is about 15ms.
-
- Posts: 22
- Joined: Wed Jan 25, 2017 8:40 pm
Re: Multithread
Bump of old thread.
Did anyone figure this out, please?
Thank you!
Did anyone figure this out, please?
Thank you!
-
- Posts: 22
- Joined: Wed Jan 25, 2017 8:40 pm
Re: Multithread
Thank you.
I tried all the way up to 32 TCP with no difference in throughput.
I tried all the way up to 32 TCP with no difference in throughput.
-
- Posts: 2458
- Joined: Mon Feb 24, 2014 11:03 am
Re: Multithread
Could you explain about experiment detail more?
-
- Posts: 22
- Joined: Wed Jan 25, 2017 8:40 pm
Re: Multithread
thisjun wrote:
> Could you explain about experiment detail more?
Yes, thank you for asking. The details can be found here:
http://www.vpnusers.com/viewtopic.php?f=7&t=7270
Thank you!
> Could you explain about experiment detail more?
Yes, thank you for asking. The details can be found here:
http://www.vpnusers.com/viewtopic.php?f=7&t=7270
Thank you!