just checked out rtpproxy from CVS and run it. It seg faulted so loaded
up gdb and found...
j = ishostseq(bindaddr[j], spa->laddr[i]) ? 0 : 1; (main.c:763)
The funny thing is j is never initialised so I'm lost how this works for
other people. Prehaps depending on the compiler settings, j gets set to 0 ?
For people who do use rtpproxy, what version are you using (I'm running
SER 0.8.14) and where did you obtain it from?
Is mediaproxy the way to go? It didn't work for me but obviously I've
made a mistake somewhere.
cheers,
Adam
Hallo,
I would like to know if the newest SER server can handle Sigcomp and TLS.
Best regards,
/Tung
--
NEU: WLAN-Router f�r 0,- EUR* - auch f�r DSL-Wechsler!
GMX DSL = superg�nstig & kabellos http://www.gmx.net/de/go/dsl
Hello Serusers,
I would like to be able to provide some users with a cpl scripting
possibility. I do this by putting them in group "cpl" .. Those who are in
that group, will get cpl_run_script and those who are not will get normal
routing.
Is it possible to check somehow if a user has a cpl script uploaded? Does
cpl_run_script return some information I can use? (script failed because
not exists or because it exited with a failure?).
I realize that I can just put the user into the group only after he has
uploaded a script, but that would require additional checking which I
wouldnt like.
You can see a mockup of a Java web-app we're building for easier cpl
creation here: http://tanesha.net:8080/cpl/ if it should have anyone's
interest ;-)
Cheers, Soren.
--
Soren Davidsen,
Univ: AUE department of computer science, room H302, +45-20676081
Home: http://tanesha.net/, +45-36947805 or +1-3602266324
Ring billigt med IP-telefoni: http://musimi.dk/, http://GratisSip.dk/
I am running SER 8.12 (with radius modules) on a Redhat Linux ES 3.0 system.
My client is MS Windows Messenger 4.72. I start MS Windows Messenger and
quickly receive the following message:
"Signing in to Communications Service failed because the service is
temporarily unavailable."
I have attached the ngrep from the server, and also a packet capture at the
client workstation running MS Win Messenger.
I am registering as morriss(a)orau.gov, and I believe the coxT(a)orau.gov is
showing up because this person is one my contact list.
Scott Morris
Enterprise Network Engineer
DOE - ORAU / ORISE
865-576-4672
I use TCP for my SIP connections. I have noticed other proxy servers
send TCP "null packet" keep-alives during a connection. What is the
feasibility of doing something like this in SER? Where would you
steer someone who would consider implementing?
In particular this would solve a firewall issue related to "inactivity
timeout". (UAC and UAS don't support Session Timers)
Jamie Yukes
Hey Girish,
Thanks for swift reply!
I had come to same conclusion about dialplan too... however, I am not
sure the best way of proceeding. In our network, most users will have
both alphanumeric usernames and 4-digit traditional style extension
numbers - both for those with 12-key phones and for easy mapping to
external numbers. For this reason, it seemed easiest to base the
Asterisk VM boxes on those 4-digit extension numbers... then rather than
add an extension for every user in *, I just have a single rule using
the {exten} variable. Otherwise I need some way of mapping alphanumeric
and digit codes to the same mailbox in *... very confusing... I thought
that it would be a lot easier to let * handle the simplest one, and make
Ser always pass it the format it's expecting... there must be way to do
a lookup(alias) and then re-write the URI with the alias before it gets
forwarded.?
Sorry if my explanation about the 404 is confusing... it's confusing for
me too! Attached is the ser.cfg (I hope it retains the indentation!)
Many thanks again,
Dave
-----Original Message-----
From: GR S [mailto:gr_sh2003@yahoo.com]
Sent: 28 July 2004 22:04
To: Dave Bath
Cc: serusers(a)lists.iptel.org
Subject: RE: [Serusers] Asterisks to ser to asterisk (voicemail)
Hello,
--- Dave Bath <dave(a)fuuz.com> wrote:
<snip>
> (1) If a user is called with their alphanumeric ID instead of their
> numerical alias, * does not pick up the call. This is as expected, as
> the dial pattern in * is _[1-9] [0-9] [0-9] [0-9]. However, it must
be
> fairly common to call people with their email addresses for example...
> so how can I make ser pass the alias to * instead of the alpha URI?
I think the problem is in your dialplan. Try adding something like this:
exten => _.,1,Dial(SIP/....
> (2) If a user is offline, I get a 404 immediately, instead of anything
> else - for example diverting immediately to vm. I don't quite
> understand this at the moment.. as I have the t_on_failure set up
before
> the location lookups... does the t_on_failure not catch 404 failures?
Sorry, I dont understand this problem. Looks to me as a misconfiguration
in the config file.
Please post your ser.cfg. There are many in this forum who can tell you
about the problem after
verifying the config file. I saw your previous post and excerpts from
the config file, but did not
quite understand that.
> Thanks again everyone,
>
> Dave
>
Best Regards,
=====
Girish Gopinath <gr_sh2003(a)yahoo.com>
__________________________________
Do you Yahoo!?
Yahoo! Mail - 50x more storage than other providers!
http://promotions.yahoo.com/new_mail
Hello Serusers.
I have some problems with how to use Mediaproxy with SEMS for voicemail and
the failure route.
I've setup a seperate Ser for SEMS voicemail to redirect from the example
and it is running ok.
In my main Ser it is working fine with redirecting to voicemail if user is
offline, but i cannot get the voicemail redirection working on the failure
route.
I'm using use_media_proxy on all invites, and have on reply processing to do
use_media_proxy also, and all calls seems to work fine.
It looks like the Mediaproxy port is closed down when on_failure is being
processed (or when the invite timer is running out?) but SEMS still use the
RTP port that Mediaproxy set to send out to and then i dont hear what's on
the other end :-S
Have anyeone used failure_route sems voicemail and mediaproxy together? an
example would be nice to read up on.
I've attached parts of my ser.cfg below:
failure_route[1] {
xlog("L_INFO","Failure route VM: ru: %ru tu: %tu fu: %fu\n");
# rewrite for voicemail.
revert_uri();
rewritehostport("212.130.58.214:5064");
append_branch();
t_relay();
break;
}
Thanks,
Jens Davidsen
Musimi.dk
Hi,
I installed (better said I unzipped, untarred and
copied files to /usr/local/...) SER.
I did not change ser.cfg or something else. I tried
to run ser by 'ser start' as it's described in "ser_howto.html".
I immidiatelly after start get message:
"Too much shared memory demanded: 33554432"
Even if I try to start ser with some arguments like debugging
they are ignored.
Problem is probably not in ser, but somewhere in my system.
Name and version of my OS:
Linux elephant 2.2.20-idepci #1 Sat Apr 20 12:45:19 EST 2002 i686 unknown
SER distribution:
ser-0.8.14_linux_i386.tar.gz
SER build:
version: 0.8.14 (i386/linux)
flags: STATS:Off, USE_IPV6, USE_TCP, DISABLE_NAGLE, DNS_IP_HACK, SHM_MEM,
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
@(#) $Id: main.c,v 1.168.4.3 2004/06/28 15:41:21 andrei Exp $
main.c compiled on 12:28:01 Jul 27 2004 with gcc 2.95
SER config:
unchanged - the same as in package
SER logs:
unavailable - SER generates nothing even in syslog there is no any message
Any idea where could be a problem?
Great thanks for any hint,
Pali - hopeless broken crying man :o)