[Olsr-dev] hanging http-requests / httpinfo+json / v9.0.3 (STILL in git)

Henning Rogge (spam-protected)
Thu Dec 17 20:09:52 CET 2015

On Thu, Dec 17, 2015 at 7:50 PM, Bastian Bittorf <(spam-protected)> wrote:
> * Ferry Huberts <(spam-protected)> [17.12.2015 18:33]:
>> The only thing I can think of that can cause this is the limit on the
>> number of outstanding responses.
>> Try raising the MAX_CLIENTS define
> ok, I inserted a lot of debug-messages and it seems
> that the real problem happens in line 383 of src/net_olsr.c
> http://www.olsr.org/git/?p=olsrd.git;a=blob;f=src/net_olsr.c;h=37cb9700be44d758c89678eefa7e59636ccda9bd;hb=bf884cfea57194e26121ec5d5223aaff9edb5b4c#l381
> if (olsr_sendto(ifp->send_socket, ifp->netbuf.buff, ifp->netbuf.pending,
> MSG_DONTROUTE, (struct sockaddr *)sin, sizeof(*sin)) <
>         0) {
>       perror(...
> when starting the daemon interactive, i can see ~1 second
> the message:
> sendto(v4) Invalid argument

funny... net_output is not called by http/txt/jsoninfo.

> now i'am struggling with formatting a good
> logmessage which gives me all the argument of
> the olsr_sendto-call - to see where the underlying prob
> is. (btw: raising MAX_CLIENTS does not help)
> as far as i can say now when query the daemon for
> http/json-output it never comes to the codepath
> of lib/info/http_headers.c -> http_header_build

I think you have a totally different problem.

do you have any special IP addresses on your mesh interfaces?


More information about the Olsr-dev mailing list