hi,
we are succesfully built an alternative for our tincd backbone with
wireguard (and on top l2tpv3/batv15).
with iperf we get up to 500Mbs on l2tpv3 level.. (upto 700 on pure
wireguard)
for optimization we need to know more and better about the used header,
the possible data-size per packet-he
Hi Jens,
I assume it was you asking in the IRC channel about this same thing
before signing out? Sorry I wasn't there when you were; I only just
now arrived home.
There actually is some optimization potential for you with regards to
the MTU. The overhead of WireGuard breaks down as follows:
- 20
Many people ask about the packet breakdown of WireGuard, and though
this is explained in [1] and [2], many find this ancient mailing list
thread, which now contains out of date information. So this email is
to bring the thread up to date, for folks who stumble upon it.
The overhead of WireGuard b
Hi,
For WireGuard overhead breakdown [1], maybe it's worth also mentioning
that N the length of encrypted data will be padded to be multiples of
16.
I am only aware of this when fragmentation was spotted. With 1500 as
MTU for ethernet, PPPoE has MTU 1492 (1500 - 8). I thought 1432 (1492
- 60) f
Hello, I hope this is the right place.
I see here[1] that if you're using IPv4 exclusively, you can get away with an
MTU of 1440. If my client only has IPv4 internet, however the server issues an
IPv6 address for use by the client, can the client still use 1440 without
fragmentation, or must th
On Thu, 17 Aug 2023 20:14:52 +
blurt_overkill...@simplelogin.com wrote:
> I see here[1] that if you're using IPv4 exclusively, you can get away with
> an MTU of 1440. If my client only has IPv4 internet, however the server
> issues an IPv6 address for use by the client, can the client still us
On Thu, 24 Aug 2023 08:50:20 -0400
Saint Michael wrote:
> This is the Achiles' heel of Wireguard. It reduces the MTU too much. Other
> tunneling techniques use a much larger MTU. I use Mikotik routers and one
> of the supported tunnels goes up to 1472. Some apps requiere a large MTU.
> Why Wiregu
On Wed, 17 Jul 2019 17:45:18 +0800
Yousong Zhou wrote:
> For WireGuard overhead breakdown [1], maybe it's worth also mentioning
> that N the length of encrypted data will be padded to be multiples of
> 16.
>
> I am only aware of this when fragmentation was spotted. With 1500 as
> MTU for ethern