On 11/04/2012 07:00 PM, Jonathan Nieder wrote: > Hi Patrik,
> I suspect there was a miscommunication here: if I understand > correctly, Ben was looking for a simple sequence of steps and, for > each step, a routing table from after that step. This would help him > to understand what was happening when you ran into this problem. > Perhaps he could run similar steps in a similar setup and compare, to > track down where the problem lies. > > See http://www.chiark.greenend.org.uk/~sgtatham/bugs.html for more > context. > > Thanks for your patience, and hope that helps. > Jonathan > Hi Jonathan, I would like to help, but the difference of what happens when it works and doesn't is seemingly none. No error messages or anything else is shown: The connection just doesn't work when I try it. It is why I suspect that if a (background) program holds a socket open, the kernel can't set the route. Although it reports the correct route. This error has also happened when I started Firefox to early and it checks for updates. When this error happens, I retry the connection (ctrl+c for openvpn and run it again). Sometimes I need to retry several (3-4) times, but usually one retry is enough. I have tested this bug with two different providers of openvpn-services and there is no difference. It has of course was easier to help the bug hunting, when the same error always happens in a sequence. Do you have any kind of dump program, to report all the status of all connections, which also reports the true state of the kernel? This error also exists in Ubuntu 11.04. (I haven't been using any later.) Best regards, Patrik -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/5096cb81.20...@gmail.com
