Hello gnugk, ser and asterisk users!
I am using asterisk as a SIP-to-H323 translator in order to call my
itsp provider that only support H.323 so this is my scenario:
--|xlite|-->|SER-SIP_Proxy|-->|Asterisk_oh323|-->|GNUGK_Proxy|-->|h323GW_MyProvider|-->PSTN
x-lite its a sip client registered in SER, SER forward all the 00+
calls to an Asterisk box which is registered as a gateway in
GNUGK(proxy mode) connected with my ITSP. everything its ok buy I have
a problem with GNUGK radius-accounting:
In ACF and CDR packets (shown below in status port 7000) I can see the
correct caller ID (the xlite number as a h323_ID) but this caller-id
is not shown in radius accounting packet, instead of that I see other
number used by asterisk to register to gnugk (Alias=1001 in
oh323.conf).
I need the correct "Calling-Station-Id" to take advantage of the gnugk
radius-acc and its prepaid support... any idea??
thanks
Rafael
GNUGK Version:
- PTLib version is 1.5.2
- OpenH323 version is 1.12.2
- Gatekeeper(GNU) Version(2.0.8)
Asterisk Version:
- Asterisk CVS version (Nov 3, 2004)
- asterisk-oh323-0.7.0.tar.gz (last version), compiled with
pwlib-v1_6_6-src.tar.gz, openh323-v1_13_5-src.tar.gz
Status Port:
------------
ACF|200.110.w.x:1720|1269_endp|14962|1411#13054224049:dialedDigits|RafoXlite:h323_ID=6606060:dialedDigits|false;
CDR|106|4e 0b 06 3d 57 2c d9 11 94 bb 00 04 76 2f 16 9a|4|Wed, 03 Nov
2004 17:42:00 -0500|Wed, 03 Nov 2004 17:42:04
-0500|200.110.w.x:1720|1269_endp|65.211.y.z:1720|1267_endp|1411#13054224049:dialedDigits|RafoXlite:h323_ID=6606060:dialedDigits|GKProxy01;
DCF|200.110.w.x|1269_endp|14962|normalDrop;
Radius ACC for the same call:
---------------------------
Wed Nov 3 17:41:57 2004
Acct-Status-Type = Start
NAS-IP-Address = 200.110.w.x
NAS-Identifier = "GKProxy01"
NAS-Port-Type = Virtual
Service-Type = Login-User
Acct-Session-Id = "03030069"
User-Name = "IP-PBX"
Framed-IP-Address = 200.110.w.x
Calling-Station-Id = "1001" <<<<------------WHY??
Called-Station-Id = "1411#13054224049"
h323-gw-id = "h323-gw-id=GKProxy01"
h323-conf-id = "h323-conf-id=760B063D 572CD911 94BB0004 762F169A"
h323-call-origin = "h323-call-origin=proxy"
h323-call-type = "h323-call-type=VoIP"
h323-setup-time = "h323-setup-time=17:41:57.000 PET Wed Nov 03 2004"
h323-remote-address = "h323-remote-address=65.211.y.z"
Acct-Delay-Time = 0
Client-IP-Address = 200.110.w.x
Acct-Unique-Session-Id = "c1746abd83fdf370"
Timestamp = 1099521717
Wed Nov 3 17:42:04 2004
Acct-Status-Type = Stop
NAS-IP-Address = 200.110.w.x
NAS-Identifier = "GKProxy01"
NAS-Port-Type = Virtual
Service-Type = Login-User
Acct-Session-Id = "03030069"
User-Name = "IP-PBX"
Framed-IP-Address = 200.110.w.x
Acct-Session-Time = 4
Calling-Station-Id = "1001"
Called-Station-Id = "1411#13054224049"
h323-gw-id = "h323-gw-id=GKProxy01"
h323-conf-id = "h323-conf-id=760B063D 572CD911 94BB0004 762F169A"
h323-call-origin = "h323-call-origin=proxy"
h323-call-type = "h323-call-type=VoIP"
h323-setup-time = "h323-setup-time=17:41:57.000 PET Wed Nov 03 2004"
h323-connect-time = "h323-connect-time=17:42:00.000 PET Wed Nov 03 2004"
h323-disconnect-time = "h323-disconnect-time=17:42:04.000 PET
Wed Nov 03 2004"
h323-disconnect-cause = "h323-disconnect-cause=10"
h323-remote-address = "h323-remote-address=65.211.y.z"
Acct-Delay-Time = 0
Client-IP-Address = 200.110.w.x
Acct-Unique-Session-Id = "c1746abd83fdf370"
Timestamp = 1099521724
--
rrgv
Hi Flynn,
Hope this is not too late for you. Initially I had a
number of issues getting this working as well. In the
end I was unable to get the stable 0_8_14 version
working with the version of rtpproxy found at
www.portaone.com. Not sure why, looking through the
list it seemed the current recommendation was to use
the unstable version from CVS, along with the rtpproxy
version found at the same location.
My first attempt to get everything running on the
unstable version failed as SER could never seem to
find rtpproxy after starting, even after specifying
the location in the ser.cfg file and ensuring the
rtpproxy.sock file had the correct permissions. I
ended up removing everything and starting over with a
new config file. The second attempt worked fine.
Here is my version info:
SER Version:
ser# ser -V
version: ser 0.8.99-dev12 (i386/freebsd)
RTPProxy Version:
ser# rtpproxy -v
20040107
OS Version:
FreeBSD 4.10
Here are the steps/commands that worked for me (as of
Oct 24th 2004):
1. Get the unstable SER version from CVS:
- set
CVSROOT=:pserver:anonymous@cvs.berlios.de:/cvsroot/ser
export CVSROOT
- cvs login
- cvs co sip_router
2. Get version of RTPProxy from CVS
- cvs co rtpproxy
3. Compile SER
- cd /root/sip_router
- gmake install
4. Compile RTPProxy
- cd /root/rtpproxy
- ./configure
- make install
5. Start RTPProxy
- /usr/local/bin/rtpproxy
6. Replace ser.cfg located in /usr/local/etc/ser with
attached file.
7. Add the following line to the new ser.cfg in the
modparam section of the file.
- modparam("nathelper","rtpproxy_sock",
"/var/run/rtpproxy.sock")
8. Set debug=3 and log_stderror=yes on your new
ser.cfg file so you can see debug
information.
9. Start SER
- /usr/local/sbin/ser
10a. This is what you want to see:
3(13842) 2(13841) rtpp_test: RTP proxy found,
support for it enabled
rtpp_test: RTP proxy found, support for it
enabled
10b. This is not what you want to see:
1(47773) ERROR: send_rtpp_command: can't read
reply from a RTP proxy
1(47773) 2(47774) WARNING: rtpp_test: can't get
version of the RTP proxy
Also, one of the tests run by the nathelper module
uses RFC1918 (private) addressing in order to
determine if an IP is behind a NAT. To simplify
troubleshooting in your test environment, only use
private addressing on devices that are actually behind
a NAT. Also set your debug level to 8 while making
calls and log the output.
The "serctl ul show" command can be used to see how
your NATed phones are registering. Here is one of my
phones located behind a NAT - 1.1.1.83 is SER and
1.1.1.89 is the NAT device. The phone IP behind the
NAT is 192.168.0.101.
...Record(0x283a6e10)...
domain: 'location'
aor : '5551234567(a)1.1.1.83'
~~~Contact(0x283a6e60)~~~
domain : 'location'
aor : '5551234567(a)1.1.1.83'
Contact :
'sip:5551234567@1.1.1.89:15060;user=phone;transport=udp'
Expires : 3592
q :
Call-ID : '1457041155(a)192.168.0.101'
CSeq : 1
replic : 0
User-Agent: 'Cisco-CP7905/1.02-040406A'
received : ''
State : CS_NEW
Flags : 1
next : 0x0
prev : 0x0
~~~/Contact~~~~
.../Record...
Regards,
Allan
--- Ahmad Faiz <flynnmachine(a)gmail.com> wrote:
> Allan,
>
> I read your post on the serusers mailing list that
> you were able to
> get SER working with rtpproxy and NAT. I've been
> trying to get this
> running too for a while now, but have been coming up
> against a wall
> every time.
>
> Would it be possible for you to share your ser.cfg
> file with me? I'm
> really desperate to get my installation up and
> running in time to meet
> the boss' deadline...
>
> thanks a bunch,
> flynn
>
______________________________________________________________________
Post your free ad now! http://personals.yahoo.ca
Hello all,
I am trying to debug a new module that I have written.
So I have set fork=no and debug=4 and commented out the
child line.
The question I have is that the how may times the child_init()
function be called ?
The child_init() function in my module sets up a socket connection
to a server and I can see that the child_init() function is getting
called 3 times. I do not understand this. Am I missing something ?
The mod_init function is getting called only once, like it is supposed to.
I have attached the output from ser startup. The module that I have written
is called gl.
Nov 2 14:07:20 sara ser: INFO: glenayre redirector(gl module) initializing
Nov 2 14:07:20 sara ser: INFO: udp_init: SO_RCVBUF is initially 65535
Nov 2 14:07:20 sara ser: INFO: udp_init: SO_RCVBUF is finally 262142
Nov 2 14:07:20 sara ser: gl:INFO: gl_child_init(): init_child #-1 / pid
<10269>
Nov 2 14:07:20 sara ser: gl:INFO: gl_child_init(): Initializing TCP
connection
Nov 2 14:07:20 sara ser: gl:INFO: gl_child_init(): TCP port = 1234
Nov 2 14:07:20 sara ser: gl:INFO: gl_child_init(): TCP host = localhost
Nov 2 14:07:20 sara ser: INFO: fifo process starting: 10270
Nov 2 14:07:20 sara ser: gl:INFO: gl_child_init(): init_child #-2 / pid
<10270>
Nov 2 14:07:20 sara ser: gl:INFO: gl_child_init(): Initializing TCP
connection
Nov 2 14:07:20 sara ser: gl:INFO: gl_child_init(): TCP port = 1234
Nov 2 14:07:20 sara ser: gl:INFO: gl_child_init(): TCP host = localhost
Nov 2 14:07:20 sara ser: SER: open_uac_fifo: fifo server up at
/tmp/ser_fifo...
Nov 2 14:07:20 sara ser: gl:INFO: gl_child_init(): init_child #1 / pid
<10268>
Nov 2 14:07:20 sara ser: gl:INFO: gl_child_init(): Initializing TCP
connection
Nov 2 14:07:20 sara ser: gl:INFO: gl_child_init(): TCP port = 1234
Nov 2 14:07:20 sara ser: gl:INFO: gl_child_init(): TCP host = localhost
any help is appreciated,
thanks,
Jignesh Gandhi
Software Engineer II
Jignesh.Gandhi(a)glenayre.com
Hi again,
i'm trying to pass all calls through ser, using ipfw on a freebsd machine
with two network interfaces.
As long as i use the real address (192.168.1.10) of ser on the client,
everything works as expected,
ser forwards all calls to the asterisk-server (10.10.10.10) on the second
network.
Using asterisk directly from the clients in the 192.168.1.0/24 network also
works.
Now i'm trying to use the direct ip of the asterisk server on the clients,
but put ser in front of it
using an ipfw forward:
ipfw add fwd 127.0.0.1,5060 udp from not 192.168.1.10 to any 5060
Basically the same as a transparent squid on a firewall.
But ser seems to completely ignore *any* request that hits him. I guess this
is because it
does not listen on the ip that is really targeted by the clients
(10.10.10.10.), can i somehow
to answer such requests also?
Thanks and regards,
Andreas
_________________________________________________________________
Listen to music online with the Xtra Broadband Channel
http://xtra.co.nz/broadband
Hi All,
I am using SJPhone application on a windows XP machine to register with the
IPTEL.org proxy server.(publicly available) In this Register request which
is sent to the server, the contact port is different from the source port.
Now the SIP invite requests coming to this machine are being sent to the
source port of the register request, instead of the contact port of the
register request. IT WAS EARLIER COMING ON THE CONTACT PORT.
Was any change done in the proxy server recently which is causing this
behavior?
Thanks
Mahesh
serusers，您好！
asterisk can not hangup .user Wildcard X100P.
when using phone call,asterisk can not hangup.
I should tpye:
soft hangup zap/1-1
then can hangup.
dev2003
dev2003(a)mail.ustc.edu.cn
2004-11-03
I have SER 0.8.14 running in 2 instances on a P4 Fedora Core 1 with 512MB RAM with SEMS for voicemail. Xten Pro softphones work perfectly.
instance one: SER -f /usr/local/etc/ser/ser.cfg -P /var/run/ser.pid
instance two: SER -f /usr/local/etc/ser/voicemail.cfg -P /var/run/servm.pid
SEMS uses the defaults.
All config files are attached.
My MVP130 operates in 3 distinct ways, based on the status of the callee.
1) The callee is offline: voicemail picks up immediately. VOIP-200410247 is the tcpdump. You see the RTP traffic come back from SER immediately, and you can hear the message on the phone, but at the word "possible" in the "call isn't available" message, the MVP130 hangs up, as though it's recieved a "BYE" message. I can't see why.
2) The Callee is online, but ignores the call (either VOIP-20041027 or VOIP-20041027-2) The phone rings forever, even past the "network timeout" setting configured on the MVP130. Traces show the RTP traffic headed back to the client, but you never hear it on the phone.
3) Callee is online, but SIP traffic fails to tranmit the NAT router, so the invite fails. This works flawlessly with the transfer to voicemail (The NAT transversal is a failure of the firewall, not NATHELPER, afaik).
Thanks in advance - all the information here is so good!
Rob
Are there any examples of how run a select query to the db from within
ser.conf , and return a value other than 0 or 1 ?
I have read in the mailing list archives that the only way to do this is to
call a shell script, is that true ?
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.777 / Virus Database: 524 - Release Date: 10/14/2004
Hi All.
I came across this posting
http://lists.iptel.org/pipermail/serusers/2004-August/010832.html where Bogdan
suggested sending REGISTER messages to multiple ser proxies using something
like this:
append_branch( dst1 );
append_branch( dst2 );
t_replicate( dst3 );
How would I best solve a problem where if I were to have say 10 ser proxies and
wanted to avoid changing the ser.cfg everytime I added/removed a new sip proxy
to the farm? I'm thinking it would be best to do this with a custom module
which would look up rows in a MySQL table. Each row would be the location of a
ser proxy which would be used in a call to append_branch(). This way I could
add new ser proxies to the farm without modifying ser.cfg on every machine and
in my ser.cfg file I could replace the above snippet with something like this:
append_proxy_locations();
Does anyone see pitfalls with this? Is there a better solution?
Regards,
Paul
__________________________________
Do you Yahoo!?
Take Yahoo! Mail with you! Get it on your mobile phone.
http://mobile.yahoo.com/maildemo
