sr-users August 2012

sr-users@lists.kamailio.org

77 participants
127 discussions

Core fm_realloc function
by Bruno Bresciani 08 Jan '13

08 Jan '13

Hi all, I observed two at 0x08173bf0 in fm_realloc (qm=0xb5f7d000, p=0xb61f0bc0, size=0) at mem/f_malloc.c:536. I don't know exactly what may be caused this core because the debug of kamailio is disabled, but looking at backtrace (gdb) and source code I could notice a relationship with TLS connection establishment (handshake). Other important fact is that core was generated by the kamalio itself with signal 6 (abort) because it can't find the pointer to memory realloc. I'd like to know if someone already observed this issue and if my analyse above is correct? Bellow is backtrace full with gdb: Core was generated by `/home2/local/kamailio/sbin/kamailio -P /var/run/kamailio.pid'. Program terminated with signal 6, Aborted. #0 0x00f30402 in __kernel_vsyscall () (gdb) (gdb) (gdb) (gdb) (gdb) bt full #0 0x00f30402 in __kernel_vsyscall () No symbol table info available. #1 0x009fec10 in raise () from /lib/libc.so.6 No symbol table info available. #2 0x00a00521 in abort () from /lib/libc.so.6 No symbol table info available. #3 0x08173bf0 in fm_realloc (qm=0xb5f7d000, p=0xb61f0bc0, size=0) at mem/f_malloc.c:536 f = (struct fm_frag *) 0xb61f0bb8 pf = <value optimized out> orig_size = 18 L=ÃbÂ¬Ã2n = (struct fm_frag *) 0xb61f0bd2 ptr = <value optimized out> hash = 23548 #4 0x0051a87f in ser_realloc (ptr=0xb61f0bc0, size=32) at ../../mem/shm_mem.h:266 No locals. #5 0x00460295 in OPENSSL_cleanse () from /lib/libcrypto.so.6 No symbol table info available. #6 0x004609f7 in CRYPTO_realloc () from /lib/libcrypto.so.6 No symbol table info available. #7 0x003efb22 in sk_insert () from /lib/libcrypto.so.6 No symbol table info available. #8 0x003efbaa in sk_push () from /lib/libcrypto.so.6 No symbol table info available. #9 0x007c2ac3 in ssl_bytes_to_cipher_list () from /lib/libssl.so.6 No symbol table info available. #10 0x007aa9f3 in ssl3_get_client_hello () from /lib/libssl.so.6 No symbol table info available. #11 0x007ab815 in ssl3_accept () from /lib/libssl.so.6 No symbol table info available. #12 0x007c1c3a in SSL_accept () from /lib/libssl.so.6 No symbol table info available. #13 0x005278e3 in tls_accept (c=0xb621af88, error=0xbf91dfc8) at tls_server.c:346 ret = <value optimized out> ssl = (SSL *) 0xb61c1878 cert = <value optimized out> tls_c = (struct tls_extra_data *) 0xb6208e58 tls_log = <value optimized out> #14 0x0052da8e in tls_read_f (c=0xb621af88, flags=0xbf91e200) at tls_server.c:1028 r = (struct tcp_req *) 0xb621affc bytes_free = 4095 bytes_read = 70 read_size = 4095 ssl_error = 0 ssl_read = 0 ssl = (SSL *) 0xb61c1878 rd_buf = "\026\003\001\000A\001\000\000=\003\001OÃ¿\vÂ¡ywP\fkÃDÃ 0/\000\a\000\005\001\000ÃÃÃÃ¯000\000\026\0009\0008\0005\000\026\000\023\000\n\0003\0002\00 202i<Ã¤<Ã¥Ã¾\031Ã¬234Â \237Ãª030Y){lÑ¶"...06Ã'Ãr9Ã¬Ã«Â¸{[Â©\t\205\037\036Ã¿HPH5Âµ\t(Ã©32Ã027Ã¦Ã²036Â¾\"\rÔ¤YÃ215uÂº\027)Ã¶000Ã±215\vÃÃ Ã©Ea\023xÃ©J\206T\204CÃ wr_buf = "\027\003\001\003Â wÂ²'\016Ñ¼Ã033ÃLÃ021Ã)Ã\220hnLÂ¦\020XÃª\236\200\tIhÃÂ¸Sf\234Ã\rÃ ?Â¯Â¼\024FÃ\200Âª/\2264RÃaÃÂ¤IpXÂ»Â¨Ã035Ã¤Qh\211Ã026Ã½4ÃU+\asrÂ£/`Â¤OÃ»1AÊ¹qUÜ¿Â·vÂ±Ã¼\205"... rd = {buf = 0xbf90dfa4 "\026\003\001", pos = 70, used = 70, size = 65536} wr = { buf = 0xbf8fdfa4 "\027\003\001\003Â wÂ²'\016Ñ¼Ã033ÃLÃ021Ã)Ã\220hnLÂ¦\020XÃª\236\200\tIhÃÂ¸Sf\234Ã\rÃ ?Â¯Â¼\024FÃ\200Âª/\2264RÃaÃÂ¤IpXÂ»Â¨Ã035Ã¤Qh\211Ã026Ã½4ÃU+\asrÂ£/`Â¤OÃ»1AÊ¹qUÜ¿Â·vÂ±Ã¼\205"..., pos = 0, used = 0, size = 65536} tls_c = (struct tls_extra_data *) 0xb6208e58 enc_rd_buf = (struct tls_rd_buf *) 0x0 n = 0 flush_flags = <value optimized out> err_src = 0x541a83 "TLS read:" x = <value optimized out> #15 0x08158ade in tcp_read_headers (c=0xb621af88, read_flags=0xbf91e200) at tcp_read.c:406 bytes = -1 remaining = <value optimized out> p = <value optimized out> r = (struct tcp_req *) 0xb621affc #16 0x08158fd4 in tcp_read_req (con=0xb621af88, bytes_read=0xbf91e204, read_flags=0xbf91e200) at tcp_read.c:871 bytes = <value optimized out> total_bytes = 0 resp = <value optimized out> size = <value optimized out> req = (struct tcp_req *) 0xb621affc dst = {send_sock = 0x1, to = {s = {sa_family = 57540, sa_data = "\221Â¿\020\000\000\000\000\000\000\000\020\000\000"}, sin = {sin_family = 57540, sin_port = 49041, sin_addr = {s_addr = 16}, sin_zero = "\000\000\000\000\020\000\000"}, sin6 = {sin6_family = 57540, sin6_port = 49041, sin6_flowinfo = 16, sin6_addr = {in6_u = {u6_addr8 = "\000\000\000\000\020\000\000\000\001\000\000\000\001\000\000", u6_addr16 = {0, 0, 16, 0, 1, 0, 1, 0}, u6_addr32 = {0, 16, 1, 1}}}, sin6_scope_id = 12}}, id = -1080958468, proto = 4 '\004', send_flags = {f = 0 '\0', blst_imask = 0 '\0'}} c = 10 '\n' #17 0x08159a0f in handle_io (fm=0x83296b0, events=1, idx=-1) at tcp_read.c:1150 ret = 4 n = 4 read_flags = 1 con = (struct tcp_connection *) 0xb621af88 s = 12 resp = <value optimized out> t = <value optimized out> #18 0x0815d653 in tcp_receive_loop (unix_sock=35) at io_wait.h:1091 No locals. ---Type <return> to continue, or q <return> to quit--- #19 0x0812885b in tcp_init_children () at tcp_main.c:4819 r = 3 reader_fd_1 = 35 pid = <value optimized out> si = <value optimized out> #20 0x080ad412 in main_loop () at main.c:1632 i = 4 pid = <value optimized out> si = (struct socket_info *) 0x0 si_desc = "udp receiver child=3 sock=192.168.166.31:5060 \000\221Â¿\002\000\000\000/\200\220\000\001\000\000\000\000\000H\000Â¨Ã¤21Â¿\001\000\000\000Â°Â¶Ã¨026Â¶\000\000\000\000\000ÃÂµ\030s2\b\002\000\000\000Ã¢f\b\000ÃÂµ\001\000\000\000\000\000\000\000\001\000\000\000\000\000\000\000Ã¿Ã¿Ã¿Ã¿Â¸Ã¤21Â¿" #21 0x080afbe7 in main (argc=3, argv=dwarf2_read_address: Corrupted DWARF expression. ) at main.c:2398 cfg_stream = (FILE *) 0x8a2d008 c = <value optimized out> r = 0 tmp = 0xbf91e674 "\212Ã¾\221Â¿\233Ã¾\221Â¿Â°Ã¾\221Â¿Â»Ã¾\221Â¿Ã\221Â¿Ã¹1Â¿\020Ã¿\221Â¿DÃ¿\221Â¿LÃ¿\221Â¿WÃ¿\221Â¿]Ã¿\221Â¿oÃ¿\221Â¿{Ã¿\221Â¿\202Ã¿\221 tmp_len = 10492229 port = 136192361 proto = -1080957480 ret = <value optimized out> seed = 21782655 rfd = 4 debug_save = <value optimized out> debug_flag = 0 dont_fork_cnt = 0 n_lst = <value optimized out> p = <value optimized out> Best Regards

4 24

[Kamailio-Users] BYE message not relayed to caller
by Vikram Ragukumar 04 Dec '12

04 Dec '12

Hello, My setup includes a cellphone that uses Wi-Fi to hookup to the internet with IP 192.168.1.101. Cellphone app, tries to register with VOIPSwitch through an outbound proxy server(Kamailio + rtpproxy), which rewrites SDP to force rtp through an rtpproxy. I can make a call from the cellphone to a PSTN number. When the PSTN endpoint hangs up, VOIPSwitch relays a BYE message to the proxy that looks like : No. Time Source Destination Protocol Info 27 18.383929 IPaddr_VoipSwitch IPaddr_proxy SIP Request: BYE sip:1234@IPaddr_VoipSwitch Frame 27 (420 bytes on wire, 420 bytes captured) Ethernet II, Src: Unispher_40:b5:39 (00:90:1a:40:b5:39), Dst: Supermic_bd:b9:bc (00:30:48:bd:b9:bc) Internet Protocol, Src: IPaddr_VoipSwitch , Dst: IPaddr_proxy User Datagram Protocol, Src Port: sip (5060), Dst Port: 7160 (7160) Session Initiation Protocol Request-Line: BYE sip:1234@IPaddr_VoipSwitch SIP/2.0 Method: BYE [Resent Packet: False] Message Header Route: <sip:IPaddr_proxy:7160;lr=on;nat=yes> CSeq: 1 BYE Via: SIP/2.0/UDP IPaddr_VoipSwitch:5060; branch=z9hG4bk180242100434182912333954 From: sip:5678@IPaddr_VoipSwitch;tag=18024210041929123187506289 Call-ID: LUOyIDm-ecy36NcUfj5G3v1lM6g5snkM To: "1234" <sip:1234@IPaddr_VoipSwitch>;tag=9XHJmiuf58vTrF7.vwflEK-G63JVPQSi Content-Length: 0 The request URI in the above message is the IP address of VOIPSwitch itself, which causes the proxy to forward the BYE message back to VOIPSwitch. The cellphone never receives the BYE message and consequently never hangs up. Any suggestions on how i can get around this problem ? Thanks and Regards, Vikram.

3 9

[OT] the role of SBCs
by Daniel-Constantin Mierla 26 Oct '12

26 Oct '12

Hello, based on the outcome of the discussion carried in the thread: http://lists.sip-router.org/pipermail/sr-users/2012-August/074480.html I am looking again the clarify some aspects out there in the VoIP world. So, as mainly dealing with proxy/sip signaling deployments, it's very often to be the first one hit by support issues claiming things don't work. Then you investigate and end up in conclusions like in the above thread: "The problem was at the SBC, __where I did not expect it__." The underlined part heats me up a bit, because I never understood from where it comes this perception that SBC is a MUST-TO-HAVE and the PERFECT (never mistaken or breaking things) node in a VoIP networks. To some extent, the SBC is just a very costly SIP ALG, and a SIP ALG is there to break the things. I don't want to start like a flame war, but is it something that I am obviously missing in regards to what benefits a SBC can bring? I see only inconveniences: - another point of failure - it is a b2bua, therefore very unlikely to offer the same performances of a proxy - if transcoding is needed, a media server can be used behind the proxy, properly protected of attacks by the proxy and eventually deployed as a farm load balanced by the proxy - if topology hinding is wanted in the b2bua fashion (not the proxy fashion with encoding headers), then the b2bua can be behind the proxy, properly protected of attacks by the proxy and eventually deployed as a farm load balanced by the proxy - nat traversal was solved long time ago in proxy environment, being scalable by deploying a farm of rtp proxy I don't want to go to other features, including the transport layer, it's a clear win of the proxy in my experience (ok, being deep involved in this project). Then, what makes the SBC so desirable in many companies/voip deployments? If any SBC user here that can share, what was the reason to buy such a device? Any conceptual functionality that cannot be achieved with the proxy as the first hop in front of the (wild) clients? Cheers, Daniel -- Daniel-Constantin Mierla - http://www.asipto.com http://twitter.com/#!/miconda - http://www.linkedin.com/in/miconda Kamailio Advanced Training, Berlin, Nov 5-8, 2012 - http://asipto.com/u/kat

10 14

is kamilio support RFC5626 now ?
by mike 16 Oct '12

16 Oct '12

dear kamilio developers : i read the kamilio 3.3.0' s new features , it say it support outbound , and i have place to handle reg-id and sip.instance . so , i wonder if the kamilio3.3.0 support RFC5626 now，so that the kamilio can be used as a edge proxy . i ask why because i can't find where kamilio add flow-token and path for the register request. thanks

5 8

kamailio core at qm_status
by Jijo 01 Oct '12

01 Oct '12

Hi All, I'm observing a core intermittently at "qm_status (qm=0x786cd000) at mem/q_malloc.c:763" for kamailio version 3.1.0 looking at the backtrace this is occurring while doing a tcp buffer overrun. Please have a look and let me know if anyone observed this issue or how can we debug it. Please find the backtrace and kamailio version osbprod-V2R0:~ # /usr/sbin/kamailio -V version: kamailio 3.1.0 (i386/linux) 21a375 flags: STATS: Off, USE_IPV6, USE_TCP, USE_TLS, TLS_HOOKS, USE_RAW_SOCKS, DISABLE_NAGLE, USE_MCAST, DNS_IP_HACK, SHM_MEM, SHM_MMAP, PKG_MALLOC, DBG_QM_MALLOC, USE_FUTEX, FAST_LOCK-ADAPTIVE_WAIT, USE_DNS_CACHE, USE_DNS_FAILOVER, USE_NAPTR, USE_DST_BLACKLIST, HAVE_RESOLV_RES ADAPTIVE_WAIT_LOOPS=1024, MAX_RECV_BUFFER_SIZE 262144, MAX_LISTEN 16, MAX_URI_SIZE 1024, BUF_SIZE 65535, PKG_SIZE 15MB poll method support: poll, epoll_lt, epoll_et, sigio_rt, select. id: 21a375 compiled on 12:38:44 Apr 26 2012 with gcc 4.5.0 #0 qm_status (qm=0x786cd000) at mem/q_malloc.c:763 #1 0x0819c0f3 in qm_debug_frag (qm=0x786cd000, f=<value optimized out>) at mem/q_malloc.c:167 #2 0x0819c352 in qm_free (qm=0x786cd000, p=0x7a6cb6ac, file=0x821cce2 "<core>: tcp_main.c", func=0x822285f "wbufq_run", line=880) at mem/q_malloc.c:457 #3 0x0815f686 in wbufq_run (tcpconn=0x7a1a7548, ev=4, fd_i=-1) at tcp_main.c:880 #4 handle_tcpconn_ev (tcpconn=0x7a1a7548, ev=4, fd_i=-1) at tcp_main.c:4141 #5 0x08169bf2 in io_wait_loop_epoll () at io_wait.h:1092 #6 tcp_main_loop () at tcp_main.c:4606 #7 0x080b0404 in main_loop () at main.c:1655 #8 0x080b1f84 in main (argc=9, argv=0xbfb4c9f4) at main.c:2446 (gdb) l 758 LOG_(DEFAULT_FACILITY, memlog, "qm_status: ", 759 "dumping free list stats :\n"); 760 for(h=0,i=0;h<QM_HASH_SIZE;h++){ 761 unused=0; 762 for (f=qm->free_hash[h].head.u.nxt_free,j=0; 763 f!=&(qm->free_hash[h].head); f=f->u.nxt_free, i++, j++){ 764 if (!FRAG_WAS_USED(f)){ 765 unused++; 766 #ifdef DBG_QM_MALLOC 767 LOG_(DEFAULT_FACILITY, memlog, "qm_status: ",

4 7

[Kamailio-Users] AudioCodes + Kamailio : Problem in SIP Message Headers
by Samuel Muller 26 Sep '12

26 Sep '12

Hello all, I recently add a classical Audiocodes Mediant 2000 with 2x 8E1, the purpose is to have several interconnections with PSTN. I configured it like this : Audiocodes registers as a gateway to the Kamailio, using a dedicated port (5062). Registration seems to be OK, and the pstn gw uses OPTIONS method to ping the proxy. I can attack the Audiocodes with a SIP phone behind Kamailio, no pbm. But the audiocodes returns some errors about SIP headers sent by Kamailio : ( sip_stack)(44732 ) AcSIPParser: Problem in SIP Message Headers [Time: 12:30:26] ( sip_stack)(44733 ) !! [ERROR] AcSIPParser: Parse Error. Unexpected symbol '0' in scheme. ALPHA expected Here you have the example of an INVITE from a SIP phone to the PSTN : ** audiocodes debug ** 4d:12h:30m:26s ( lgr_flow)(44730 ) ---- Incoming SIP Message from 77.246.81.132:5060 ---- INVITE sip:0323719001@77.246.81.136:5062;transport=udp SIP/2.0 Record-Route: <sip:77.246.81.132;lr=on;ftag=71078b346a20fb3eo0;nat=yes> Via: SIP/2.0/UDP 77.246.81.132;branch=z9hG4bKdace.1ab1d59.0 Via: SIP/2.0/UDP 192.168.0.113:5060;rport=15170;received=77.246.81.162;branch=z9hG4bK-b432f96 From: "Sam" <sip:0123451010@sip.720.fr <sip%3A0123451010(a)sip.720.fr>>;tag=71078b346a20fb3eo0 To: <sip:0323719001@sip.720.fr <sip%3A0323719001(a)sip.720.fr>> Call-ID: 944d8aec-27503ee6(a)192.168.0.113 CSeq: 102 INVITE Max-Forwards: 49 Contact: "Sam" <sip:0123451010@77.246.81.162:15170> Expires: 240 User-Agent: Linksys/SPA941-5.1.8 Content-Length: 281 Allow: ACK, BYE, CANCEL, INFO, INVITE, NOTIFY, OPTIONS, PRACK, REFER Supported: 100rel, replaces Content-Type: application/sdp P-Asserted-Identity: <0123451010> Remote-Party-ID: <0123451010>;party=caller;privacy=none;screen=yes v=0 o=- 26933860 26933860 IN IP4 192.168.0.113 s=- c=IN IP4 77.246.81.133 t=0 0 m=audio 35038 RTP/AVP 18 0 8 101 a=rtpmap:18 G729a/8000 a=rtpmap:0 PCMU/8000 a=rtpmap:8 PCMA/8000 a=rtpmap:101 telephone-event/8000 a=fmtp:101 0-15 a=ptime:30 a=sendrecv a=nortpproxy:yes ( sip_stack)(44732 ) AcSIPParser: Problem in SIP Message Headers [Time: 12:30:26] ( sip_stack)(44733 ) !! [ERROR] AcSIPParser: Parse Error. Unexpected symbol '0' in scheme. ALPHA expected ( sip_stack)(44734 ) !! [ERROR] Message type: INVITE [Time: 12:30:26] ( sip_stack)(44735 ) !! [ERROR] Source header: [Time: 12:30:26] ( sip_stack)(44736 ) !! [ERROR] Line: 17. Column: 23 [Time: 12:30:26] The outgoing INVITE from Kamailio is exactly the same received by the AudioCodes. When I searched over Google, I just found 2 answers about Asterisk / Audiocodes unsolved problem, but no more informations. I supposed that the problem is as indicated : " s=- " where source is empty in place of "NULL" / "0" or something like this ... Someone can confirm or already met the problem ? Many thanks all :) .Sam.

5 5

[Kamailio-Users] Having problems using RTPProxy to bridge internal/external networks
by Joe Hart 15 Sep '12

15 Sep '12

Hi all, For a project on which I'm currently working, I am having some problems figuring out how to correctly configure Kamailio to communicate with RTP Proxy in order to send media into and out of a network with private IP address ranges. I have a proxy set up to send the SIP traffic, and all of this is working fine. However, I'm having some trouble getting the RTP Proxy set up. Currently, when the call is connected, the offer/answer is made and RTP Proxy seems to be taking over, but I'm having trouble getting my audio to flow in both directions. Examination of the traffic coming into and out of this machine seems to indicate that the IP addresses aren't being mangled correctly. Specifically, it appears the internal IP address isn't being changed to reflect the IP address of the machine on which RTP Proxy is running, so that when the caller tries to send audio back, the IP it's given to reply to is 10.10.x.x, which obviously won't work. I have tried experimenting with specifically setting IP addresses in the rtpproxy_offer() and _answer() methods to no avail, as well as setting various flags in those methods. However, I must admit that I'm not entirely sure what's happening under the hood with these methods, or what rtpproxy is doing with that information when it gets it. Rather than continue to hack at this by trial and error, I'm hoping someone here can point me in the right direction. Any advice, example code or pep talks would be greatly appreciated. Thanks in advance, -- Joe Hart Voice Systems Integrator Evariste Systems Web : http://www.evaristesys.com/ Tel : (+1) (678) 954-0670 Direct : (+1) (678) 954-0673

8 37

Rtpproxy re-INVITE handling
by Spencer Thomason 11 Sep '12

11 Sep '12

Hello, I'm using Kamailio 3.2.4 for NAT traversal using rtpproxy_manage() in a largely stock script. Everything works great until the far end (on a public ip) sends a t.38 re-INVITE. The 200OK from the NATed UAC then doesn't trigger rtpproxy and the private IP in the sdp causes the fax to fail. Any help handling these re-INVITEs would be greatly appreciated. I'm happy to post traces if that helps describe the situation. The network topology looks like this: NATed UAC -> Kamailio on a public IP -> UAS on a public IP Thanks in advance, Spencer Connected by DROID on Verizon Wireless

2 12

Possible bug in authentication
by martian＠centrum.sk 07 Sep '12

07 Sep '12

Hello to everybody. I am currently working with Kamailio 3.3.1 on RedHat. The "loose_route" function was not working correctly and I observed some very strange behaviour (not as one described in the documentation of the function). I have found that there needs to be a port included in the "alias" variable for the loose_route function to work correctly. However, upon adding the port to alias, the INVITE messages were no longer authenticated (Kamailio just accepted them and didn't send proxy-auth header in 407 message). My alias: alias="domain.ch:5060" Examining default routing logic, I found the problem here: if (is_method("REGISTER") || from_uri==myself) { # authenticate requests ... } The "from_uri==myself" was no longer evaluated as true, because there was a port at the end of the alias. The FROM Header of the INVITE messages looks like: From: "acc1" ;tag=12345 ..so .. no port number there. Btw, I have fixed this with replacing the "myself" list with my own defined variable MY_DOMAIN. #!define MY_DOMAIN ".*(a)domain.ch" So now the condition looks like this: if (is_method("REGISTER") || from_uri=~MY_DOMAIN) { ... } I am not sure if this is a bug that needs to be fixed or not. I am just pointing my finger at it and I hope it will contribute to the development. Also, a valid description of this behavior (when using port in alias) would be appreciated. Cheers, Martin

3 8

FW: Kamailio as sipRedirect
by Fatima Chahrour~Vanrise Support 07 Sep '12

07 Sep '12

Dears, I have an installed 3CX where I want to generate a call from it to the Kamailo to get a list of suppliers to terminate the call to. Thus, the Kamailio should act as a sip redirect to the 3CX (for routing decision). Have you tried such configuration before? And how do I configure the 3CX on the Kamailio? Regards, F Charour

3 12

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

sr-users August 2012