I have a http proxy across tinc linked remote networks. It works generally well, but few times a day it gives a timeout, which disconnects some more sensitive clients. It is nginx based and setting longer timeouts, disabling or extending buffers did not help. Do you have any experience with a tinc or lower level networking parameters which could be used to mitigate such an annoying behaviour? Thanks, Serge
I actually run my setup in pretty much the same way ... VPS w/ Nginx <- Tinc -> Dedicated physical server with multiple sites. Though I have never had problems with it. Could you provide some more detail about what you mean by timeout? Are clients getting a timeout connecting to Nginx or is Nginx giving an error connecting to the remote sites? Do you have any other type of logging, like a continuous ping from the Nginx box to the Remote box? On Tue, Nov 13, 2018 at 1:54 AM Sergiusz Pawlowicz <sergiusz at pawlowicz.name> wrote:> I have a http proxy across tinc linked remote networks. > > It works generally well, but few times a day it gives a timeout, which > disconnects some more sensitive clients. It is nginx based and setting > longer timeouts, disabling or extending buffers did not help. > > Do you have any experience with a tinc or lower level networking > parameters which could be used to mitigate such an annoying behaviour? > > Thanks, > Serge > _______________________________________________ > tinc mailing list > tinc at tinc-vpn.org > https://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/20181113/776a041d/attachment.html>
On Tue, 13 Nov 2018 at 17:00, CaffeineAddiction <caffeineaddiction at gmail.com> wrote:> > I actually run my setup in pretty much the same way ... VPS w/ Nginx <- Tinc -> Dedicated physical server with multiple sites. Though I have never had problems with it. Could you provide some more detail about what you mean by timeout? Are clients getting a timeout connecting to Nginx or is Nginx giving an error connecting to the remote sites? Do you have any other type of logging, like a continuous ping from the Nginx box to the Remote box?The service which is proxied is Matrix (https://github.com/matrix-org/synapse), which is probably more specific setup than you have. It is what I see in the nginx reverse proxy error log: 2018/11/13 09:47:25 [error] 12191#12191: *5005629 connect() failed (110: Connection timed out) while connecting to upstream, client: xxx.xxx.xxx.xxx Please note, if the Matrix is on the same server as nginx, there are no timeouts - so definitely I am sure it is related to tinc managed WAN hop. Serge