Hi,
Currently I'm constantly killing batmand and starting it up again (different debugging levels, etc).
Sometimes I end up with a 2nd gate tunnel (so I'd have a gate0 and a gate1).
I understand (guess?) that it's because I do kill -9 and don't give batmand a chance to clean up.
Normally is it true that there would only ever be one tunnel (gate0) ?
when multiple nodes are deployed is this still always the case?
thanks,
Derek
Hey,
On Sat, Nov 22, 2008 at 02:27:45PM -0000, Derek C wrote:
Hi,
Currently I'm constantly killing batmand and starting it up again (different debugging levels, etc).
Sometimes I end up with a 2nd gate tunnel (so I'd have a gate0 and a gate1).
I understand (guess?) that it's because I do kill -9 and don't give batmand a chance to clean up.
Yep, you're probably too fast for the kernel to clean up. With kill -9 you don't give batman the chance to clean it up. Shouldn't a normal "terminate" (without -9) be enough? We catch this signal, clean up and exit properly in this case.
Normally is it true that there would only ever be one tunnel (gate0) ?
Yup, only one.
when multiple nodes are deployed is this still always the case?
Multiple nodes is no problem (multiple gateways).
The only way to have multiple gateX is to start multiple different batman (which is not possible under normal conditions), or killing/restarting it too fast. :)
regards, Simon
On Saturday 22 November 2008 22:27:45 Derek C wrote:
Currently I'm constantly killing batmand and starting it up again (different debugging levels, etc).
Sometimes I end up with a 2nd gate tunnel (so I'd have a gate0 and a gate1).
I understand (guess?) that it's because I do kill -9 and don't give batmand a chance to clean up.
Normally is it true that there would only ever be one tunnel (gate0) ?
when multiple nodes are deployed is this still always the case?
Yes, you always will have only one gate0. If you happen to have gate1 make sure that you firewalling (NAT) is ok. Actually, I have an idea how to make that additional NAT rule unnecessary but I can't test it. If you have time join our IRC channel and we can discuss it.
Why are you killing batman with SIGKILL anyways ? Do you want fast access to the debug output ? You can start batman with the "-d <debug_level>" option to prevent batman from forking into the background (CTRL + C will kill batman).
Regards, Marek
Hi Marek,
It was pure lazyness that I was using kill -9 :)
Derek
On Sat, November 22, 2008 3:44 pm, Marek Lindner wrote:
On Saturday 22 November 2008 22:27:45 Derek C wrote:
Currently I'm constantly killing batmand and starting it up again (different debugging levels, etc).
Sometimes I end up with a 2nd gate tunnel (so I'd have a gate0 and a gate1).
I understand (guess?) that it's because I do kill -9 and don't give batmand a chance to clean up.
Normally is it true that there would only ever be one tunnel (gate0) ?
when multiple nodes are deployed is this still always the case?
Yes, you always will have only one gate0. If you happen to have gate1 make sure that you firewalling (NAT) is ok. Actually, I have an idea how to make that additional NAT rule unnecessary but I can't test it. If you have time join our IRC channel and we can discuss it.
Why are you killing batman with SIGKILL anyways ? Do you want fast access to the debug output ? You can start batman with the "-d <debug_level>" option to prevent batman from forking into the background (CTRL + C will kill batman).
Regards, Marek
B.A.T.M.A.N mailing list B.A.T.M.A.N@open-mesh.net https://list.open-mesh.net/mm/listinfo/b.a.t.m.a.n
b.a.t.m.a.n@lists.open-mesh.org
-
Derek C -
Marek Lindner -
Simon Wunderlich