“A picture stolen from the west to attract attention”
In our past articles, we talked about how to work with VDS on Windows Server Core 2019 at our new UltraLight tariff for 99 rubles per month. We offer another way to use this tariff. This time we’ll talk about what’s better to choose if you need a VPN for the lazy or a static IP address, which by the way is more convenient to use instead of Hamachi and everything else, if you really want to play heroes or Warcraft 3 on a local network. It's not about tuning, let's talk about performance.
Testing methodology
RRAS and SoftEther were chosen based on ease of installation, support for the L2TP protocol and the ability to control via GUi.
For SoftEther and RRAS, a L2TP connection with a shared key was used through standard Windows tools. As it was established, it was tested.
The operating system for SoftEther is Ubuntu 18.04 LTS, for RRAS Windows Server Core 2019. All operating systems before the tests received the latest updates on the state from 11/21/2019.
On the second-generation Hyper-V virtual machine, 1 GB of RAM was installed, as well as processor limits. The order of the test groups is as follows:
For all 8 cores:
- With no restrictions
- 50% limit
- 25% limit
- 5% limit
- 1% limit
On 4 cores:
- With no restrictions
- 50% limit
- 25% limit
- 5% limit
- 1% limit
Per core:
- With no restrictions
- 50% limit
- 25% limit
- 5% limit
- 1% limit
All VPN servers applied the settings out of the box and NAT was enabled. All virtual machines are on the same host and on the same virtual switch.
To evaluate the network, testing was performed between the server and the client without connecting via VPN.
The test was conducted using the TamoSoft Throughput Test in TCP only mode, for the tables and graphs the values “ave” were taken. Data was collected for 5 minutes 30 seconds for each test.
For a better understanding of the limits of both implementations, we first test the throughput of the virtual switch.
This is what the results in the testing program looked like. Further, all results will be wrapped in tables.
As you can see, the virtual switch is not a bottleneck in testing and almost gives a theoretical limit of 10 gigabits.
How the “physical” test network looked
Results:
Per core:
In discipline, both servers run flush to the same core.
On 4 cores:
On 8 cores:
Here we clearly see which solution scales best with the number of cores. By reducing the performance of each of the cores, RRAS compensated for the losses by their number, which SoftEther did not.
System memory consumption
The amount of RAM consumed by SoftEther increased from the number of cores, from 122 to 177 MB, but still less than RRAS.
The RRAS service itself weighs about 200 megabytes in memory, minus the total system consumption.
Bandwidth under different conditions
Total throughput without any processor restrictions.
If you still have not chosen the solution that is right for you, perhaps this table will help you make your choice. The total throughput in CPU shortage mode is shown.
Please note that on four and one core SoftEther'a performance is higher than eight. Such low performance is not found anywhere, but testing itself indicates how the algorithm really scales well on the number of cores.
Conclusion:
Connecting to SoftEther when the processor was limited did not pass the first time, I had to first increase the limit, connect and then lower the limit, this imposes a restriction on its installation in very thin environments. RRAS always logged in instantly.
If you have a machine with a lot of cores, prefer RRAS. And for SoftEther, you can leave 4 cores. The author, if he had used it, would have left only one core for it.
What and where to place - decide for yourself. If you have 99 rubles for VPS with Windows Server on board, RRAS will still be the best choice.
