I'm trying to trace a complete dialog on a loadbalancer/proxy:
UAC <-> Proxy <-> backend
109.235.34.a <-> 109.235.32.b <-> 185.61.68.c
Proxy runs 4.3.x (but with 4.4.x I appear to have the same results).
siptrace config:
=============================================
modparam("siptrace", "duplicate_uri", "sip:10.11.12.13:9060")
modparam("siptrace", "hep_mode_on", 1)
modparam("siptrace", "trace_to_database", 0)
modparam("siptrace", "trace_flag", 22)
modparam("siptrace", "trace_on", 1)
=============================================
After REQINIT (which does pike and handles OPTIONS), I setflag(22) the
message.
=============================================
route(REQINIT);
setflag(22);
route(NATDETECT);
route(WITHINDLG);
=============================================
Next depending in where the message comes from:
=============================================
if(!is_in_subnet($si, backends))
{
ds_select_dst(1, 0);
t_on_failure("RTF_DISPATCH");
}
else
{
remove_hf("Route");
loose_route();
}
route(RELAY);
=============================================
sip_trace() is being called in the routes:
WITHINDLG
MANAGE_BRANCH
MANAGE_REPLY
MANAGE_FAILURE
RTF_DISPATCH
The resulting duplicate packets are:
11:20:21.037 INVITE 109.235.34.a:39621 109.235.32.b:5060 tcp
11:20:21.040 INVITE 109.235.34.a:39621 109.235.32.b:5060 tcp
11:20:21.047 401 185.61.68.c:5060 109.235.32.b:5060 udp
11:20:21.073 ACK 109.235.34.a:39621 109.235.32.b:5060 tcp
11:20:21.117 INVITE 109.235.34.a:39621 109.235.32.b:5060 tcp
11:20:21.135 100 185.61.68.c:5060 109.235.32.b:5060 udp
11:20:21.450 180 185.61.68.c:5060 109.235.32.b:5060 udp
11:20:27.929 200 185.61.68.c:5060 109.235.32.b:5060 udp
11:20:28.106 ACK 109.235.34.a:39621 109.235.32.b:5060 tcp
11:20:30.625 BYE 185.61.68.c:5060 109.235.32.b:5060 udp
11:20:30.659 BYE 185.61.68.c:5060 109.235.32.b:5060 udp
11:20:30.663 BYE 109.235.32.b:5060 109.235.34.a:39621 tcp
11:20:30.664 200 109.235.34.a 39621 109.235.32.b:5060 tcp
11:20:30.664 200 109.235.32.b:5060 185.61.68.c:5060 udp
What is missing are the forwarded replies, there is an ACK that matches
the 401, but I have no clue what message was send to the UAC. Adding a
sip_trace after REQINIT shows nothing extra.
How (if possible) do I get all messages being send via the proxy?
I could use sngrep/sipgrep, but that is an other process to monitor for
failure.
Hi guys,
I have the following multipart body
Content-Type: multipart/mixed;boundary=uniqueBoundary
Remote-Party-ID: <tel:1112222;phone-context=+54>
Content-Length: 364
--uniqueBoundary
Content-Type: application/sdp
v=0
o=user1 53655765 2353687637 IN IP4 111.111.111.111
s=-
c=IN IP4 111.111.111.111
t=0 0
m=audio 6000 RTP/AVP 8
*a=rtpmap:8 PCMA/8000*
*--uniqueBoundary--*
Content-Type: application/gtd
Content-Disposition: signal;handling=optional
IAM,
CPC,20
GCI,885sdfasdf1123
--uniqueBoundary--
As you can see *--uniqueBoundary *before *a=rtpmap:8 PCMA/8000* has not an
extra '\r\n'. Is this correct? Or should have it?
The thing is that rtpengine is receiving sdp with no final '\r'n' from
kamailio:
Dump for 'offer' from 127.0.0.1:37388: { "sdp": "v=0#015#012o=user1
53655765 2353687637 IN IP4 111.111.111.111#015#012s=-#015#012c=IN IP4
111.111.111.111#015#012t=0 0#015#012m=audio 6000 RTP/AVP 8#015#012*a=rtpmap:8
PCMA/8000"*, "ICE": "remove", "direction": [ "pub", "priv" ], "replace": [
"origin", "session-connection" ], "transport-protocol": "RTP/AVP",
"call-id": "1-17755(a)111.111.111.111", "received-from": [ "IP4",
"111.111.111.111" ], "from-tag": "1", "command": "offer" }
And the resulting sdp from rtpengine is this:
--uniqueBoundary
Content-Type: application/sdp
v=0
o=user1 53655765 2353687637 IN IP4 172.16.213.16
s=-
c=IN IP4 172.16.213.16
t=0 0
m=audio 35066 RTP/AVP 8
*a=rtpmap:8 PCMA/8000a=sendrecv*
a=rtcp:35067
The problem is related with that '\r\n' but Is this a malformed message or
a possible bug?
Thanks in advance!
Diego
Hi,
I was wondering whether there is some kind of documentation of all
functions currently exported for KEMI by Kamailio modules. I know I
can find them somehow via grep, but I could not find a page in the
Kamailio wiki listing all of them. Is there any?
Best Regards,
Sebastian
[I see I posted this to -dev instead of -users; fixing that -JimC]
I'm working on converting some configs to kemi lua and have a couple of
questions:
First:
Does KSR.pv.get() support multiple pseudos in a single call?
The examples such as misc/examples/kemi/kamailio-basic-kemi-lua.lua
make several calls, but also do things like KSR.pv.get("$T_branch_idx]")
to avoid having the append "]" to the string. That makes me wonder why
that file has code like this:
KSR.dbg("new branch [".. KSR.pv.get("$T_branch_idx]")
.. " to ".. KSR.pv.get("$ru") .. "\n");
instead of like:
KSR.pv.get("[$T_branch_idx] to $ru\n")
Second:
Is it still the case that, eg, event_route[xhttp:request] cannot be in
the kemi lua code?
There was another. Perhaps I'll recall what it was... :(
-JimC
--
James Cloos <cloos(a)jhcloos.com> OpenPGP: 0x997A9F17ED7DAEA6
Good afternoon!
Can you please tell us how the Kamailio works with shared memory. The fact
is that there was an accident in which Kamailio reported an error - little
common memory. This happened a day later as I added the dialplan module.
After that, I began to monitor the values:
used_size
real_used_size
max_used_size
free
And I saw that I can not properly assess.
I note that the server does not process calls, it is idle. After the server
is started, the indicators are as follows:
used_size: 63
real_used_size: 146
max_used_size: 146
free 106
After rebooting the module (for example, drouting, in which the data has not
changed):
used_size: 63
real_used_size: 193
max_used_size: 231
free 62
I found, a description, the reasons for the increase in memory after
rebooting the modules, I quote "As suggested by Daniel, for carrierroute
reload, we need the same amount of memory as it used to be. Block when
reloading the table. "
But why:
1. after increasing the memory the server frees only real_used_size
2. That that is specified as real_used_size on my observations can not be
released in current of several days (probably weeks).
3.If in this state again to reboot the module (the same one that led to an
increase in memory), the memory does not increase.
Question: during what time should kamailio release the memory specified as
real_used_size? Also what is considered as real_used_size?
--
View this message in context: http://sip-router.1086192.n5.nabble.com/Kamailio-428-memory-usage-tp161327.…
Sent from the Users mailing list archive at Nabble.com.
Hi,
Is there any possibilities of "on the fly" option in kamailio.cfg.
For Example:
I need to change my carrier IP address in the lcr_gw table, kamailio.cfg
should takes it on the fly. Without restarting the kamailio.
Thanks & Kind Regards,
Logeshwaran G
