Hi Duarte,
Not sure if I understand it correctly but are you trying to set up an HA
pair of dispatchers?
If so I am doing it differently in that, I use the Virtual/floating IP
(VIP) as socket. This way the master node (which has got the VIP) can send
OPTIONs pings out fine but the backup node won't be able to send.
Functionally it is fine, but the backup node keeps logging errors because
it doesn't have the VIP:
ERROR: <core> [core/udp_server.c:599]: udp_send(): sendto(sock, buf:
0x7f0e6b54e2e0, len: 372, 0, dst: (192.168.213.143:5060), tolen: 16) -
err: Network is unreachable (101)
ERROR: tm [../../core/forward.h:231]: msg_send_buffer(): udp_send failed
When FO happens, the backup node becomes master and takes the VIP, and it
functions correctly. But I haven't worked out how to avoid the ERROR
loggings on the backup node. Maybe someone else has any better
ideas/suggestions?
Cheers,
Yufei
Hello,
I'm trying to get Kamailio to retry INVITE.
We have a Freeswitch which is sending INVITE to a sip user via Kamailio
proxy using static ip/location.
All is well, calls are flowing, but sometimes INVITE from kamailio to user
is lost and no attempt to retry is occurring and so after 30 sec kamailio
just returns 408 Request timeout.
Call flow is like this:
FS Kamailio User
--- INVITE -->
<-- 100 Trying
--- INVITE --->
(Waiting 30 secs)
<-- 408 Timeout
--- ACK --->
So assume we have failed in handling retries somehow. Kamailio is running
stateless, but other retries are working, so I assume this is not a problem?
Can someone perhaps point us in the right direction on where to solve this?
Kamailio is version 5.2.4
Thank you...
/Allan
Hi,
Like Olle said RTPEngine is perfectly capable of handling one leg in RTP
the other in SRTP. When looking up RTPEngine's documentation, may be worth
checking these out:
SRTP
SDES-unencrypted_srtp
SDES-unencrypted_srtcp
SDES-unauthenticated_srtp
e.g. try apply these flags to rtpengin_offer() on the incoming RTP leg, to
get RTPEngine to add encryption to the outgoing leg etc.
Cheers,
Yufei
Hi Andrew,
Like Alex said, have a look at the commands RTPEngine received and the
responses it sent in the history of the call, and see if RTPEngine has ever
received any commands regarding this call, or, if the call has been deleted
from RTPEngine. You can raise the RTPEngine logging level to 6 for example
so you can see the command/responses of RTPEngine.
Cheers,
Yufei
I didn't find any mention in registrar module README that save()
function saves also branch flags and that lookup() retrieves them.
Is that documented somewhere?
-- Juha
Hi there,
Every now and then, the value of active dialogs metric spikes to an
erroneous huge number, messing up my otherwise beautiful grafana gauges :(
Example:
kamailio_dialog_active_dialogs 18446744073709551614 1602812239115
kamailio_dialog_early_dialogs 0 1602812239115
kamailio_dialog_expired_dialogs 2 1602812239115
kamailio_dialog_failed_dialogs 7361 1602812239115
Anybody faced the same symptoms, any clues what might be causing this?
Thanks,
--Sergiu
Hi List,
I am using Topos module to hide network topology. I am having a problem of
routing the ACK request for REINVITE. Here is the schema:
A KAMAILIO B
1. --------- INVITE -------->
2. <---- 180 RINGING ---
3. <---- 200 OK ---
4. ---- ACK --->
5. <--------- (RE) INVITE --
6. --------- OK --->
7. <---- ACK ----- (THIS ACK IS SENT TO
PRIVATE IP OF A)
I use TOPOS module along with fix_nated_contact as I would do normally for
natted UAs. The first ACK (4) is routed correctly but not the second one
(7), which is routed to the PRIVATE IP of A instead of routing to its
public address.
Has anyone run into this problem while using TOPOS module? How could I make
them work together?
Best Regards,
Minh Phan
Hi Team,
Pls guide me on following point.
:- By using kamailio (P2P call), want to play hold on music during hold
call. Shall it be possible without any external media server like
(RTPProxy/media proxy).
Regards
Amit Pal
Hi,
I have followed below link to setup rtp engine.
https://open5gs.org/open5gs/docs/tutorial/02-VoLTE-setup/
RTP engine is started without any errors and it is enabled in pcscf also. I
am using the config files of pcscf as per the above link .. still rtp
packets are not reaching pcscf from UE . What could be the problem?
The only difference is that i have deployed ims in kubernetes environment
as pods.
Kindly help.
Thanks,
Pavithra
Hey all,
I've been using the PHONENUM module, which uses Google's libphonenumber
library, to perform number validation among other handy things within the
route script. One problem I've run into however is the metadata in the
latest stable (5.4) repo is pretty stale.
I'm curious if anyone is running this module, and if so how they've gone
about updating the metadata used by the underlying library. If I have to
ditch the repo and compile everything manually, that's not a huge deal.
Just hoping there's another way.
Thanks!
Zach
