Response time on Tinc VPNs, Bandwith on Tinc-VPN

[Donald Pearson]

The "10 megabit" connection that you see is not accurate.  The TAP-Win32
adapter will run at hardware and network speed.  I believe the reported link
speed is just residue from the days when virtual adapters didn't exist.
Personally, I've never created a VPN over fast ethernet, but on two
endpoints with 20 megabit internet connections, I have seen 17.5 megabits of
VPN traffic.  Well above the reported 10 megabit link.

Also, "TCPOnly = yes" is a Host configuration
variable<http://www.tinc-vpn.org/documentation/tinc_4.html#Host-configuration-variables>.
You have it placed as a main configuration
variable<http://www.tinc-vpn.org/documentation/tinc_4.html#Main-configuration-variables>
where
I don't believe it is doing anything.

You have "mode = switch" in your main configuration file.  I therefore think
the "Subnet=" variables in your host configuration files may be irrelevant.

On Thu, Dec 17, 2009 at 8:30 AM, Guus Sliepen <guus at tinc-vpn.org> wrote:

> On Thu, Dec 17, 2009 at 12:21:25AM +0100, Johann Stieger wrote:
>
> > response time: how is it possible to increase the response time on Tinc
> VPNs.
> > When I use tinc over a fast or giga-bit network connection the
> ping-response
> > time is normally about 1800 to 2000 ms. Sometimes I get ping response
> times
> > at about 2.500 to 3.000 ms over a normal Ethernet Connection or a
> > television-cable connection within the same providers-network in the same
> > city. The ping-time on the underlaying network connection is about 10 to
> 30
> > ms. So when I use Windows Remote Desktop it is terrible when I have to
> wait 2
> > or 3 seconds to wait for a response of the chance of the desktop.
> [...]
> > I use 3 Windows XP and one Windows Vista Notebooks as client and a
> Windows Small Business Server 2008 as server. On the client-side the
> config-file looks like this:
> [...]
> > Mode=switch
> > TCPOnly=yes
>
> Try running tinc without TCPOnly = yes. If you force TCP, then all traffic
> will
> go through a single TCP connection, which can indeed significantly decrease
> performance. The reason is explained here:
>
> http://sites.inka.de/~W1011/devel/tcp-tcp.html
>
> If you are not already using tinc 1.0.11, please upgrade. Also, add the
> following to the host config files:
>
> PMTUDiscovery = yes
>
> This will let tinc automatically determine the maximum size for UDP packets
> without fragmentation between nodes, and if UDP traffic is impossible, it
> will
> automatically fall back to TCP.
>
> --
> Met vriendelijke groet / with kind regards,
>     Guus Sliepen <guus at tinc-vpn.org>
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.10 (GNU/Linux)
>
> iEYEARECAAYFAksqMoIACgkQAxLow12M2nv1bgCfYWAz+Jb1Boa0bdI1QRu3GZgB
> sYMAoJxa4qb7Adsr2TDHA157Qs8N6gbx
> =P+wN
> -----END PGP SIGNATURE-----
>
> _______________________________________________
> tinc mailing list
> tinc at tinc-vpn.org
> http://www.tinc-vpn.org/cgi-bin/mailman/listinfo/tinc
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.tinc-vpn.org/pipermail/tinc/attachments/20091217/f3be2f01/attachment.htm>