[OpenSIPS-Users] dns srv lookups happening during relay

[Bogdan-Andrei Iancu]

Hi Tito,

Yes, your SIP URI does not have any protocol or port indication, so 
NAPTR lookup is perform.

Now, of the call flow you describe : DNS based discovery is to be done 
for initial request only - the sequential requests (in dialog) should 
simply follow the path discovered by the initial request. So, better 
post somewhere the pcap or so.

Regards,

Bogdan-Andrei Iancu
OpenSIPS Founder and Developer
http://www.opensips-solutions.com

On 15.08.2015 00:41, Tito Cumpen wrote:
> I enabled debug level 6 and I am noticing this
> : DBG:core:mk_proxy: doing DNS lookup...
> Aug 14 20:42:09 cloud-server-06 /sbin/opensips[22814]: 
> DBG:core:sip_resolvehost: no port, no proto -> do NAPTR lookup!
>
> Here it claims that this request is lacking a port and and a proto on 
> this branch I am appending the branch like so
>
>         avp_pushto("$du","sip:sisterproxyIP");
>
>         append_branch();
>
>
> the sister proxy is where the user resides. This is actually confusing 
> the remote opensips  proxy(same version) because now it sends all in 
> dialog request via a tcp socket all thought the 200 was sent via udp 
> leg. 2 requests came into the sister proxy. I'm not sure why the one 
> gets a 404 and the proceeds the only difference is the udp request has 
> two rr headers. It appears that the remote proxy might not have 
> cancelled the dialog even thought it sent a negative reply. I have a 
> trace of this if needs to be supplied.
>
>
> Thanks,
>
> Tito
>
>
>
> On Fri, Aug 14, 2015 at 3:52 PM, Tito Cumpen <tito at xsvoce.com 
> <mailto:tito at xsvoce.com>> wrote:
>
>     Group,
>
>
>     I am using :
>
>     version: opensips 2.2-dev (x86_64/linux)
>
>     flags: STATS: On, DISABLE_NAGLE, USE_MCAST, SHM_MMAP, PKG_MALLOC,
>     F_MALLOC, FAST_LOCK-ADAPTIVE_WAIT
>
>     ADAPTIVE_WAIT_LOOPS=1024, MAX_RECV_BUFFER_SIZE 262144, MAX_LISTEN
>     16, MAX_URI_SIZE 1024, BUF_SIZE 65535
>
>     poll method support: poll, epoll_lt, epoll_et, sigio_rt, select.
>
>     git revision: b7db080
>
>     main.c compiled on 22:43:41 Jun 25 2015 with gcc 4.8.3
>
>
>
>     I have a strange issue in which dns SRV lookups and routing is
>     happening although I have set the RURI domain as an alias in my
>     config. On some calls this will  eventually screw up the dialog
>     termination once a bye is sent because it makes an attempt to
>     transmit the bye to this SRV entry.
>
>     I see this error in the logs.
>
>     Aug 14 18:36:25 cloud-server-06 /sbin/opensips[15805]:
>     INFO:core:init_sock_keepalive: TCP keepalive enabled on socket 47
>     Aug 14 18:36:25 cloud-server-06 /sbin/opensips[15805]:
>     ERROR:core:tcpconn_async_connect: poll error: flags 1c
>     Aug 14 18:36:25 cloud-server-06 /sbin/opensips[15805]:
>     ERROR:core:tcpconn_async_connect: failed to retrieve SO_ERROR
>     [server=192.XX.XX.XX:38798] (111) Connection refused
>     Aug 14 18:36:25 cloud-server-06 /sbin/opensips[15805]:
>     ERROR:core:proto_tcp_send: async TCP connect failed
>     Aug 14 18:36:25 cloud-server-06 /sbin/opensips[15805]:
>     ERROR:tm:msg_send: send() for proto 2 failed
>
>     Aug 14 18:36:25 cloud-server-06 /sbin/opensips[15805]:
>     ERROR:tm:t_forward_nonack: sending request failed
>
>
>
>
>
>
> _______________________________________________
> Users mailing list
> Users at lists.opensips.org
> http://lists.opensips.org/cgi-bin/mailman/listinfo/users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensips.org/pipermail/users/attachments/20150817/8df72a2b/attachment.htm>