Wireguard openvpn performance useless info @ wiki

[fow0ryl]

I'm searching on google for some informations about wireguard performance compared to openvpn.
Found https://openwrt.org/docs/guide-user/services/vpn/wireguard/performance
and this in openwrt wiki. https://openwrt.org/docs/guide-user/services/vpn/openvpn/performance

For me these pages are useless as useless can be.
There are no informations how these values come about?
Which encryption is used by openvpn?

The openvpn throuput on my "old" XEN Server with 2 XEONs is much higher than 500Mbits. And the throuput of my historic x86 with 8088 CPU will never reach 150Mbit.

In essence this diagrams are useless. And the pages are not new at all. So I vote to delete or hide this pages in their current state. Or at least add a fat hint.

I think the best way to visualize may be to add missing values/devices to https://openwrt.org/toh/views/toh_vpn_performance?datasrt=openvpn%20performance
and generate a diagram on the fly.

Henning

[trendy]

Comparative Throughput Testing Including NAT, SQM, WireGuard, and OpenVPN

Overview While there's been a lot of individuals testing various routers, the question still always seems to come up, "What can I expect from Router X?" I won't say that my testing is going to definitively answer those questions, but at least it should be a self-consistent data set. To be very clear, this is a benchmark. Like most benchmarks, the performance you get in your own environment may be different. If you have a craptastic ISP, a lossy ISP connection, a flakey cable, hosts pummeling y…

[fow0ryl]

Of course. "What can I expect from Router X?" was my question too.
But the wiki page gives no hint, that the values are a from a personal benchmark.
And nothing about the circumstances.
I would have no problem if all of this would be described on the page.
But there is only a diagram without any necessary background information.
Just useless in this form ...

If someone like Mr. Trump finds this page randomly via google search and he looks on the page his conclusion could be: Oh my god, openwrt has very low quality. Never ever use open source software....

So less pages with more understandable and reliable info are better then even more pages.

I know Mr. Trump doesn't care about truth....

[tmomas]

Those diagrams are not intended to give absolute numbers to three decimals, but only as a rough orientation for users who are asking "Which device should I get if I want to achieve 100Mbit wireguard performance?".

Since the performance doesn't vary that much between encryption algorithms (someone wrote +/-5%), the max number provided is good enough for the purpose.

@fow0ryl

I think the best way to visualize may be to add missing values/devices to https://openwrt.org/toh/views/toh_vpn_performance?datasrt=openvpn%20performance
and generate a diagram on the fly.

Yes, that would indeed be the ideal way, however, we are missing users who enter those values in the dataentries.

[faser]

@tmomas

Yes, that would indeed be the ideal way, however, we are missing users who enter those values in the dataentries.

I guess @fow0ryl just volunteered to setup some test routers doing openvpn to wireguard comparisons to provide those detailed numbers so that in future no one need to complain anymore about those pages not giving detailed information

[fow0ryl]

For now I have done only some openssl tests, with various devices...
If you compare the results, you will see that the used chiper has much more impact as a few percent.

I was able to nearly double the throughput when changing from AES256 to ChaChaCha with openvpn. Tested with a friend of mine. He has build a self-compiled openvpn 2.5 developer version on Manjaro-x86-64-Linux some time ago. But that's not the topic here.

And again. I have no problems with incomplete pages. But the shown info should be understandable. A link to background information may have been ok.

I know: Most developers hate to write documentation. But a beginner is not able to write complex docs, because he did not know what he is talking about. So it's not my intention to belitte the work. But I'm a beginner in openwrt and I see that there is much room for improvement for the good of all.
Is it really better do spend time here to explain things again and again? Or is it better to spend time to improve the docs and reduce repetetive questions?

[faser]

@fow0ryl

Is it really better do spend time here to explain things again and again? Or is it better to spend time to improve the docs and reduce repetetive questions?

I agree with this point but Openvpn/Wireguard is not specific to openwrt and the lower CPU requirement of Wireguard has been investigated more then once and I guess there are enough such reports out there on the internet.