On Sat, Nov 19, 2011 at 11:43:40 +0100, Filippo Sallemi wrote:
Hi all, I'm working on a simple script that check if node is connected. The scope of this script is make every node able to autorescue in case of change channel on some gateway node. To do this check if i'm not a gateway try to ping my gateway ipaddress then if it not respond try to batctl ping my gateway mac and than if not respond for a number of tries the node try to find another channell with the same bssid. The problem is that batct ping every time exit with status 0 istead of behaviour of ping command that exit with 1 if ping fail. At this time I have to use some regex to check if batctl ping status command. It is possible to fix batctl ping to work such as ping command?
I think it is, we just need the proper patch :D By the way, I see that the ping command returns SUCCESS even if only one echo_reply is received. Therefore if you send 5 requests and get only 1 reply (because the node somehow changed channel while you were pinging it), your script will tell you that the node is UP anyway. Is this the wanted behaviour?
However a meaningful return value for batctl could be a good idea..(not for ping only)
Rgds
-- Filippo Sallemi