[Swan] Windows 10 client to libreswan VPN server: CHILD SA: no local proposal matches remote proposals

Mirsad Goran Todorovac mirsad.todorovac at alu.unizg.hr
Sun Jan 9 21:48:57 EET 2022


On 1/9/2022 8:42 PM, Paul Wouters wrote:

> On Sun, 9 Jan 2022, Mirsad Goran Todorovac wrote:
>
>>>  That's because most likely your l2tp layer went through userland 
>>> xl2tpd.
>>>  it can be configured to use kernel l2tp.ko but that usually has 
>>> issues.
>>
>> I have tried to deploy kernel mode L2TP, but I failed. What I get 
>> from xl2tpd is:
>
> xl2tpd has not been well maintained in the last 10 years. They have
> repeatedly broken things and accepted untested patches that solve one
> user's problem, without having a test suite.
>
> I wouldn't attempt to try xl2tpd anymore. And there are no other better
> l2tp daemons either - openl2tp also had issues.
I'm sorry to hear this. For correctness sake, I sort of wanted to have 
this working :(
>> I think I could write a paper on this comparison if I manage to get 
>> both protocols IKEv1 and IKEv2 running under same conditions?
>
> The comparison can really just say "l2tp kernel supposedly faster but
> has been abandoned and sees no real use cases anymore and does not work
> out of the box".

I guess it could, as I would stress the benefits of using IKEv2 together 
with the availability of kernel mode encryption (which is also the default).

I understand that there is no real interest in debugging xl2tpd as IKEv2 
is now preferred.

Thank you for the clarification, so I waste no more time trying to 
figure it out :)

Kind regards,
Mirsad

--
Mirsad Goran Todorovac
CARNet sistem inženjer
Grafički fakultet | Akademija likovnih umjetnosti
Sveučilište u Zagrebu
-- 
CARNet system engineer
Faculty of Graphic Arts | Academy of Fine Arts
University of Zagreb, Republic of Croatia
tel. +385 (0)1 3711 451
mob. +385 91 57 88 355



More information about the Swan mailing list