strange tinc error with many nodes

[Guus Sliepen]

On Tue, Dec 13, 2005 at 06:40:12PM +0100, Kai Köhler wrote:

> we currently set up a large tinc network with 2 central Nodes (these
> nodes connecting to each other).
[...]
> If the count of satellites reaches 40, many strange error log entries
> are visible (but only in the log file of the 2nd central box). if we
> lower the number of satellites that connecting to these two boxes the
> errors gone.
>  
> These error occurs an some satellites (less 3)
> 
> Dec 13 19:17:20 gt-fw-nfp-konstanz-scala tinc.nfp_kn_scala[2044]:
> Sending meta data to nfp_c_vpn (1.2.3.4 port 30665) failed: Resource
> temporarily unavailable Dec 13 19:17:20 gt-fw-nfp-konstanz-scala last
> message repeated 7 times
[...]
> The connection most times work. Tinc is 1.0.4 or 1.0.3 no difference
> between.
> 
> There are reasons for these interruptions? Do we something wrong?

Probably because of small TCP buffers or slow network connections. Try
adding "BlockingTCP = yes" to all host config files.

-- 
Met vriendelijke groet / with kind regards,
    Guus Sliepen <guus at sliepen.eu.org>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : http://brouwer.uvt.nl/pipermail/tinc/attachments/20051214/1ede42dc/attachment.pgp