Hi,
I am using MTREE and DIALPLAN modules to load lots of info to kamailio. (6
million rows).
When kamailio was running with 3.2.1 (no mem_join=1 option), the used size
was increasing but the process of loading the data was fast eanough.
I upgraded to 3.3.2 and set mem_join=1. Now the loading process take about
10 time longer and sometimes stops kamailio from responding to traffic.
Any ideas?
Thanks,
Uri
Hi
I am trying to run the Ubuntu nightlies from
http://deb.kamailio.org/kamailiodev-nightly and loading the p-cscf config
files.
Things were working fine while I was on the 20130110 nightlies, but when I
upgraded to 20130122 I get the following errors. Any have an idea why
sem_post is undefined and how to fix this?
0(1980) ERROR: <core> [sr_module.c:572]: ERROR: load_module: could not
open module </usr/lib64/kamailio/modules/ims_registrar_pcscf.so>:
/usr/lib64/kamailio/modules/ims_registrar_pcscf.so: undefined symbol:
sem_post
0(1980) ERROR: <core> [sr_module.c:572]: ERROR: load_module: could not
open module </usr/lib64/kamailio/modules/cdp.so>:
/usr/lib64/kamailio/modules/cdp.so: undefined symbol: sem_post
0(1980) ERROR: <core> [sr_module.c:572]: ERROR: load_module: could not
open module </usr/lib64/kamailio/modules/ims_qos.so>:
/usr/lib64/kamailio/modules/ims_qos.so: undefined symbol: sem_post
Thanks
-Barry Flanagan
Hello List.
The last weekend our kamailio process crashed with this error :
Jun 24 06:55:08 pxh /usr/local/sbin/kamailio[10661]: : <core>
[pass_fd.c:293]: ERROR: receive_fd: EOF on 29
Jun 24 06:55:08 pxh /usr/local/sbin/kamailio[10542]: ALERT: <core>
[main.c:751]: child process 10601 exited by a signal 11
Jun 24 06:55:08 pxh /usr/local/sbin/kamailio[10542]: ALERT: <core>
[main.c:754]: core was generated
Jun 24 06:55:08 pxh /usr/local/sbin/kamailio[10542]: INFO: <core>
[main.c:766]: INFO: terminating due to SIGCHLD
Can someone tell me what seems to be the problem here?
We’re using :
version: kamailio 3.2.2 (x86_64/linux) 98ba92-dirty
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, DEFAULT PKG_SIZE 4MB
poll method support: poll, epoll_lt, epoll_et, sigio_rt, select.
id: 98ba92 -dirty
compiled on 12:45:36 Feb 2 2012 with gcc 4.4.6
And this is the bt full from gdb
Core was generated by `/usr/local/sbin/kamailio -m 1024'.
Program terminated with signal 11, Segmentation fault.
#0 mi_profile_list (cmd_tree=<value optimized out>, param=<value optimized
out>) at dlg_profile.c:789
789 if ( mi_print_dlg( rpl,
ph->dlg, 0)!=0 )
Missing separate debuginfos, use: debuginfo-install
filesystem-2.4.30-2.1.el6.x86_64 glibc-2.12-1.25.el6_1.3.x86_64
keyutils-libs-1.4-1.el6.x86_64 krb5-libs-1.9-9.el6_1.2.x86_64
libcom_err-1.41.12-7.el6.x86_64 libselinux-2.0.94-5.el6.x86_64
mysql-libs-5.1.52-1.el6_0.1.x86_64 nss-softokn-freebl-3.12.7-1.1.el6.x86_64
openssl-1.0.0-20.el6_2.1.x86_64 pcre-7.8-3.1.el6.x86_64
zlib-1.2.3-26.el6.x86_64
(gdb) bt full
#0 mi_profile_list (cmd_tree=<value optimized out>, param=<value optimized
out>) at dlg_profile.c:789
node = <value optimized out>
rpl_tree = 0x20a6790
rpl = 0x20a67b0
profile = 0x7fe99a2dda00
ph = 0x0
profile_name = <value optimized out>
value = <value optimized out>
i = <value optimized out>
#1 0x00007fe9dce287b1 in run_mi_cmd (fifo_stream=<value optimized out>) at
../../lib/kmi/mi.h:77
No locals.
#2 mi_fifo_server (fifo_stream=<value optimized out>) at fifo_fnc.c:509
mi_cmd = 0x208f8f0
mi_rpl = <value optimized out>
hdl = 0x0
line_len = 41
file_sep = 0x2039017 ""
command = 0x7fe9dfdbdef9 "profile_list_dlgs"
file = 0x7fe9dfd98d78 "/tmp/openser_receiver_27809"
f = 0x2039018
reply_stream = 0x20b0980
#3 0x00007fe9dce2a640 in fifo_process (rank=<value optimized out>) at
mi_fifo.c:247
fifo_stream = 0x209a0b0
#4 0x00007fe9dce2a9de in mi_child_init (rank=0) at mi_fifo.c:211
pid = <value optimized out>
#5 0x00000000004dba41 in init_mod_child (m=0x7fe9dfc6ba10, rank=0) at
sr_module.c:886
No locals.
#6 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc6bd08, rank=0) at
sr_module.c:883
No locals.
#7 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc6c248, rank=0) at
sr_module.c:883
No locals.
#8 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc6c748, rank=0) at
sr_module.c:883
No locals.
#9 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc6d288, rank=0) at
sr_module.c:883
No locals.
#10 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc6d928, rank=0) at
sr_module.c:883
No locals.
#11 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc6dee0, rank=0) at
sr_module.c:883
No locals.
#12 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc6e1f0, rank=0) at
sr_module.c:883
No locals.
#13 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc6e5d0, rank=0) at
sr_module.c:883
No locals.
#14 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc6e830, rank=0) at
sr_module.c:883
No locals.
#15 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc6ebd0, rank=0) at
sr_module.c:883
No locals.
#16 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc6f220, rank=0) at
sr_module.c:883
No locals.
#17 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc6f648, rank=0) at
sr_module.c:883
No locals.
#18 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc6fa60, rank=0) at
sr_module.c:883
No locals.
#19 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc6fe40, rank=0) at
sr_module.c:883
No locals.
#20 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc702c0, rank=0) at
sr_module.c:883
No locals.
#21 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc70ae8, rank=0) at
sr_module.c:883
No locals.
---Type <return> to continue, or q <return> to quit---
#22 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc70e18, rank=0) at
sr_module.c:883
No locals.
#23 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc71100, rank=0) at
sr_module.c:883
No locals.
#24 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc716c0, rank=0) at
sr_module.c:883
No locals.
#25 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc71df0, rank=0) at
sr_module.c:883
No locals.
#26 0x00000000004db9c4 in init_mod_child (m=0x7fe9dfc723c0, rank=0) at
sr_module.c:883
No locals.
#27 0x000000000046077d in main_loop () at main.c:1661
i = <value optimized out>
pid = <value optimized out>
si = 0x0
si_desc = "udp receiver child=15 sock=64.76.154.35:5060", '\000'
<repeats 20 times>, "\003", '\000' <repeats 15 times>,
"\030\325\a\232\351\177\000\000\001\000\000\000\002", '\000' <repeats 11
times>,
"\001\000\000\000\000\000\000\000\003\000\000\000\000\000\000\000)\000\000\000\000\000\000"
#28 0x0000000000461f13 in main (argc=<value optimized out>,
argv=0x7fff2f2b2848) at main.c:2475
cfg_stream = <value optimized out>
c = <value optimized out>
r = <value optimized out>
tmp = 0x7fff2f2b3f76 ""
tmp_len = 0
options = 0x5a3178
":f:cm:M:dVhEb:l:L:n:vrRDTN:W:w:t:u:g:P:G:SQ:O:a:A:"
ret = -1
seed = 1850310455
rfd = <value optimized out>
debug_save = <value optimized out>
debug_flag = <value optimized out>
dont_fork_cnt = <value optimized out>
n_lst = <value optimized out>
p = <value optimized out>
(gdb)
Regards,
Ricardo.-
Dear Sir,
I am using kamailio 1.5.0, having installed and configured as given on the
kamailio site
<http://www.kamailio.org/dokuwiki/doku.php/install:kamailio-1.5.x-from-svn>
on my PC having Centos operating system. I have configured the config files
kamailio.cfg, kamctlrc accordingly as given on site.
when I start the kamailio using kamctl start command. it gives the error
message like
*[root@localhost ~]# kamctl start
database engine 'MYSQL' loaded
Control engine 'FIFO' loaded
INFO: Starting Kamailio :
ERROR: PID file /var/run/kamailio.pid does not exist -- Kamailio start
failed
*please* *help me out to solve this problem and suggest me where I am
wrong. * *I have configured kamalio so many times it gives me the same kind
of error message.
*
*
-
With Regards,
Dinesh Kumar Gautam
Noida,
+919911838275
While I have seen other posts I have not found anything that describes my scenario.
Kamailio sets in front of a group of asterisk servers which are used in a round robin dispatch group. All asterisk servers share a common database in which voicemail boxes are defined and messages are stored.
I'm looking for a way to make message waiting work in this example. I have tried creating a view with basic peer information for asterisk to use for message notification, however asterisk does not 'see' these peers until a call is made to the specific user and the peer info is retrieved from the database, which can delay notifications severely rendering them useless. I also want to avoid sending a copy of the registrations to asterisk servers, as the purpose of the distributed system is to eliminate the large number of sip registrations asterisk needs to manage.
I have also seen a couple of script methods that look at the voicemail directory structure for file changes and trigger an application like sipsak to generate the notifications and let kamailio relay them. From what I see, the consensus is not to go this route. In my case the script would need to look at the database level, and maybe some sort of database trigger could be used.
What are others doing / what do others recommend as the best way to handle message waiting with a situation where kamailio sets in front of a voicemail server?
Any input is appreciated.
Dan-
Hello,
I followed the step by step guide (http://kb.asipto.com/asterisk:realtime:kamailio-3.3.x-asterisk-10.7.0-astdb) that describe the realtime integration between Kamailio and Asterisk. I have no problem with registration but when I try a call from 101 to 102 I get the followng error:
[Mar 31 01:18:44] NOTICE[32330][C-00000006]: chan_sip.c:25195 handle_request_invite: Call from '101' (192.168.1.100:5060) to extension '103' rejected because extension not found in context 'DEFAULT NULL'.
Kamailio and Asterisk are running in the same machine.
Any idea about the cause of this problem?
Best Regards,
Theo
Hi,
I have an older installation of Kamailio (3.1.1).
It is configured with
modparam("usrloc", "db_mode", 2)
I tried to change this to
modparam("usrloc", "db_mode", 3)
to not have to restart kamailio when I change the alias table directly from
SQL.
But then kamailio shutsdown after about one second with this error:
Mar 26 20:38:55 sip-3-1 /usr/local/sbin/kamailio[11510]: ERROR: <core>
[db.c:408]: invalid version 0 for table trusted found, expected 5 (check
table structure and table "version")
Mar 26 20:38:55 sip-3-1 /usr/local/sbin/kamailio[11510]: ERROR: permissions
[trusted.c:250]: error during table version check.
Mar 26 20:38:55 sip-3-1 /usr/local/sbin/kamailio[11510]: ERROR: <core>
[sr_module.c:832]: init_mod_child(): Error while initializing module
permissions (/usr/local/lib/kamailio/modules_k/permissions.so)
Mar 26 20:38:55 sip-3-1 /usr/local/sbin/kamailio[11510]: ERROR: <core>
[pt.c:481]: ERROR: fork_tcp_process(): init_child failed for process 28,
pid 11510, "tcp receiver child=6"
Mar 26 20:38:55 sip-3-1 /usr/local/sbin/kamailio[11510]: ERROR: <core>
[tcp_main.c:4811]: ERROR: tcp_main: fork failed: Success
Mar 26 20:38:55 sip-3-1 /usr/local/sbin/kamailio[11512]: : <core>
[pass_fd.c:293]: ERROR: receive_fd: EOF on 44
Mar 26 20:38:55 sip-3-1 /usr/local/sbin/kamailio[11482]: ALERT: <core>
[main.c:738]: child process 11510 exited normally, status=255
Mar 26 20:38:55 sip-3-1 /usr/local/sbin/kamailio[11482]: INFO: <core>
[main.c:756]: INFO: terminating due to SIGCHLD
I have not made any changes to the permissions/trusted module, so I suspect
it is some shared database conection problem.
Is it a known bug in this version? I would prefer not to upgrade the
installation and hopes one of you can point me to a patch I can backport.
--
Morten Isaksen
For immediate release:
ATLANTA, GA (1 April 2013)--Evariste Systems LLC, an Atlanta-based
consultancy specialising in Kamailio-based VoIP infrastructure solutions
for the ITSP and CLEC market, has announced that beginning in the second
quarter of 2013, it will be abandoning its Kamailio-based technology
portfolio to focus on its new role as a preferred VAR (Value Added
Reseller) for Acme Packet (NASDAQ:APKT).
"It is with a heavy heart that we abandon five years of Kamailio-oriented
work and the Canonical SIP Routing Platform product derived from it,"
said Alex Balashov, the principal of the company.
"However, the reality is that investment in open-source VoIP technology
is a dead end. From a technological point of view, we have lagged very
badly in meeting the needs of today's sophisticated VoIP market, and it's
time to cut our losses. Asterisk, Kamailio, FreeSWITCH--all this stuff
just hasn't kept up with the pace of evolution of 3GPP, ETSI, and ITU
standards. We are tired of saying 'sorry, we don't support IMS or
H.323' to our resultingly dwindling customer base. Does anyone
actually run an all-SIP network?"
Starting in early April, Evariste will begin providing value-added
consultancy related to the implementation of the Acme Packet Net-Net
Session Director. In Balashov's view, "the Net-Net SD is the only
product capable of meeting the perimeter security, routing and peering
needs of today's VoIP service delivery environment."
Fred Posner, the director of Team Forrest, a Palner Group integration
and consultancy operation based in the Jacksonville, Florida area,
agreed:
"SIP is a tiny piece of the telephony puzzle. The big boys of
ClueCon [an interoperator revenue-sharing consortium] want DIAMETER-based
interdomain peering policy control, H.323, MGCP, and IMS. IMS is pretty
much how VoIP architecture is done now. We got out of the Asterisk
business just in time, right before Mitel swallowed the PBX world.
I'm glad to see Evariste is finally seeing the light, and I'm sure its
shareholders are too."
Posner also believes Evariste's lack of support for TDM interfaces
accounted for dwindling market share.
"Have you seen CSRP? It's SIP in, SIP out. Real inter-LATA haulers
and application service providers use TDM and leave SIP for things
like voicemail. I can't plug my DS3s into a SIP proxy, so I just
don't think there was any real demand for the sort of thing they
were doing."
Noting Oracle's US$2.5bn acquisition of Acme Packet in early February,
as well as its more recently announced buyout of Tekelec, a Siris
Capital Group portfolio company, Balashov remarked: "The obvious
shift to an Oracle-centric telephony paradigm was a kind of validation,
if you will, of our decision to unload our dead weight and sign on
to the revolution in unified communications."
Sean McCord, of CyCORE Systems, an Atlanta-based software consulting
house and long-time Evariste creditor, agreed that there was a natural
synergy between Evariste's shift to Acme Packet and Oracle's dominance
of telephony infrastructure.
"Oracle is a forward-thinking telecom pioneer," McCord said.
"The telephone is Oracle, and Oracle is the telephone."
Balashov also noted that a tightening regulatory environment and new
consumer protection rules helped hasten the decision to embrace the
more professionalised Acme Packet product portfolio.
John Knight, Senior Engineer at Hendersonville, NC-based Ringfree
Communications, one of Evariste's oldest channel partners, said:
"As one of Evariste's long-time disties, we were jittery about exposure
to CALEA and the QA requirements of large call centers. We tried to
make do, but at some point we just had to put the relationship on
stop. I'm all in favour of open, but there's just no open-source
software out there that does call recording, and that's the bottom line
for us. In the end, we had to restructure some debt just to get
bondholders to let us source a proprietary solution on tick."
In a thematically related move, Evariste will be dropping its heavy
use of the open-source PostgreSQL database manager for its rating and
reporting tools.
"The business case for standardising on Oracle's databases could not be
clearer. With Oracle Database 11g's support of warehousing and OLTP,
the real mystery is why we didn't go there sooner," said Balashov.
Carlos Alvarez, a director at Televolve, a growing Phoenix-area VoIP
operator, recently spearheaded a move away from Evariste's PostgreSQL-
based call detail record (CDR) storage solution to one running atop
Microsoft SQL Server 2008.
Alvarez commented: "Evariste had a nice idea, in a cute, David-and-Goliath
kind of way, but we're processing over five hundred phone calls a day
now. Are we really going to store those kinds of volumes in an
open-source database? Might as well just put it all in flat text
files at that point. Phone service is an uptime game. You can't
compromise on this stuff. What if someone needs to call 911?"
Asked to summarise his expectations, Balashov said: "I hope this turns us
around in a big way. We were wrong to think that nobody cared about
stuff like P-CSCFs, or that you could deliver even rudimentary VoIP
to the premise without the expansive feature set of a comprehensive
solution like the Net-Net SBC. I can only hope the market forgives us
for betting on 'SIP Express Router' and its ilk back in the day, and
gives us a chance to do it right in round two."
Fred Posner, of Team Forrest, added: "Besides, if you look at the Git
repository, Kamailio hasn't had any code contributions in at least five
years. It seems everyone's figured out this pure SIP stuff is defunct
and hokey."