17 Oct
2023
17 Oct
'23
7:30 p.m.
Hi all,
I am using Kamailio (5.7.2, Debian 11, Python KEMI) in a
mutli-homed environment with the topos module to hide the topology.
I have noticed that when I use `set_send_socket` or `$fsn` to force the
socket it's sent from, it breaks topos and the ACK is not proxied to the
other leg of the call. I am calling record_route() *after* forcing the
socket.
If I disable TOPOS the ACK works as expected, and the signalling looks
correct.
Or, with TOPOS enabled, if I comment out set_send_socket TOPOS works as
expected.
If I disable enable_double_rr, TOPOS works as expected and the ACK is
forwarded, but without the double RR subsequent in-dialog requests use the
wrong socket.
One other observation, when I change the send socket I also see these
warnings in the logs:
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1348]: dlg_onroute():
tight matching failed for ACK with
callid='!!:MByLMlFAM.NfWxFAM.cAMxyfWjyLz.yAO.y6MxF1MxVZWG4ZMy**'/55,
ftag='2023101714101800015'/19,
ttag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36 and direction=0
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1355]: dlg_onroute():
dialog identification elements are
callid='2023101714101800015@2900-0601-0284-80'/37, caller
tag='2023101714101800015'/19, callee
tag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36
Is anyone using TOPOS + forcing the socket, or could someone advise where I
am going wrong?
Thanks in advance
Matthew
18 Oct
18 Oct
12:18 a.m.
Hi all,
I've dug into this a bit more. Firstly I enabled debug logs and spotted the
following record-route header being loaded from redis:
21(28) DEBUG: PY3 {ACK}: topos_redis [topos_redis_storage.c:1079]:
tps_redis_load_dialog(): r[5]: s[<sip:172.24.0.10:5070
;sn=internal;r2=on;lr;ftag=2023101722345500002;did=c1f.18b1;nat=yes>,<sip:127.0.0.8;line=sr-N6IAzBFlWJZLWxP7WBN7z.VXN6ZQNUKJoSIBzwVLHRQ7z6fLz6g43RNQMByLMlFAM.NLMBM4W.jAMxyAMB1qCRPQ3ltEOBFZ3BFXoEt4H9IINA**>]
127.0.0.8 is the wrong IP which explains why the ACK was not being
forwarded correctly. A quick look in the source shows it's related to topoh.
I had modparam("topoh", "mask_callid", 1) in my config so I disabled
it and
sure enough the ACK worked as expected which gets me a bit closer to
finding the issue.
Looking at the docs for topos and topoh it looks like things have changed
since I used it last and I should be using the following instead:
modparam("topoh", "use_mode", 1)
modparam("topos", "mask_callid", 1)
But with those configured things go really wrong:
18(25) CRITICAL: PY3 {INVITE}: <core> [core/mem/q_malloc.c:501]: qm_free():
BUG: bad pointer 0x7ffc73e3ca90 (out of memory block!) called from core:
core/data_lump.c: free_lump(470) - ignoring
5(11) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
5(11) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
5(11) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
5(11) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
5(11) ERROR: <core> [core/parser/parse_fline.c:271]: parse_first_line():
parse_first_line: bad message (offset: 0)
5(11) ERROR: <core> [core/parser/msg_parser.c:748]: parse_msg(): ERROR:
parse_msg: message=<>
5(11) ERROR: <core> [core/receive.c:376]: receive_msg(): core parsing of
SIP message failed (172.24.0.21:5070/1)
3(9) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
3(9) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
3(9) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
3(9) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
Does anyone have any ideas what's going on here?
Thanks
Matthew
On Tue, Oct 17, 2023 at 6:30 PM Marrold <kamailio(a)marrold.co.uk> wrote:
Hi all,
I am using Kamailio (5.7.2, Debian 11, Python KEMI) in a
mutli-homed environment with the topos module to hide the topology.
I have noticed that when I use `set_send_socket` or `$fsn` to force the
socket it's sent from, it breaks topos and the ACK is not proxied to the
other leg of the call. I am calling record_route() *after* forcing the
socket.
If I disable TOPOS the ACK works as expected, and the signalling looks
correct.
Or, with TOPOS enabled, if I comment out set_send_socket TOPOS works as
expected.
If I disable enable_double_rr, TOPOS works as expected and the ACK is
forwarded, but without the double RR subsequent in-dialog requests use the
wrong socket.
One other observation, when I change the send socket I also see these
warnings in the logs:
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1348]: dlg_onroute():
tight matching failed for ACK with
callid='!!:MByLMlFAM.NfWxFAM.cAMxyfWjyLz.yAO.y6MxF1MxVZWG4ZMy**'/55,
ftag='2023101714101800015'/19,
ttag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36 and direction=0
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1355]: dlg_onroute():
dialog identification elements are
callid='2023101714101800015@2900-0601-0284-80'/37, caller
tag='2023101714101800015'/19, callee
tag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36
Is anyone using TOPOS + forcing the socket, or could someone advise where
I am going wrong?
Thanks in advance
Matthew
9:03 a.m.
In the header of the topic you talking about topos, but inside the messages
you talking about topoh.
They are 2 different modules. If you usr them together - they can
conflictin case how they affect message. Try use or just topoh, or just
topos.
On Wed, 18 Oct 2023, 00:45 Marrold via sr-users, <
sr-users(a)lists.kamailio.org> wrote:
Hi all,
I've dug into this a bit more. Firstly I enabled debug logs and spotted
the following record-route header being loaded from redis:
21(28) DEBUG: PY3 {ACK}: topos_redis [topos_redis_storage.c:1079]:
tps_redis_load_dialog(): r[5]: s[<sip:172.24.0.10:5070
;sn=internal;r2=on;lr;ftag=2023101722345500002;did=c1f.18b1;nat=yes>,<sip:127.0.0.8;line=sr-N6IAzBFlWJZLWxP7WBN7z.VXN6ZQNUKJoSIBzwVLHRQ7z6fLz6g43RNQMByLMlFAM.NLMBM4W.jAMxyAMB1qCRPQ3ltEOBFZ3BFXoEt4H9IINA**>]
127.0.0.8 is the wrong IP which explains why the ACK was not being
forwarded correctly. A quick look in the source shows it's related to topoh.
I had modparam("topoh", "mask_callid", 1) in my config so I disabled
it
and sure enough the ACK worked as expected which gets me a bit closer to
finding the issue.
Looking at the docs for topos and topoh it looks like things have changed
since I used it last and I should be using the following instead:
modparam("topoh", "use_mode", 1)
modparam("topos", "mask_callid", 1)
But with those configured things go really wrong:
18(25) CRITICAL: PY3 {INVITE}: <core> [core/mem/q_malloc.c:501]:
qm_free(): BUG: bad pointer 0x7ffc73e3ca90 (out of memory block!) called
from core: core/data_lump.c: free_lump(470) - ignoring
5(11) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
5(11) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
5(11) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
5(11) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
5(11) ERROR: <core> [core/parser/parse_fline.c:271]: parse_first_line():
parse_first_line: bad message (offset: 0)
5(11) ERROR: <core> [core/parser/msg_parser.c:748]: parse_msg(): ERROR:
parse_msg: message=<>
5(11) ERROR: <core> [core/receive.c:376]: receive_msg(): core parsing of
SIP message failed (172.24.0.21:5070/1)
3(9) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
3(9) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
3(9) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
3(9) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
Does anyone have any ideas what's going on here?
Thanks
Matthew
On Tue, Oct 17, 2023 at 6:30 PM Marrold <kamailio(a)marrold.co.uk> wrote:
Hi all,
I am using Kamailio (5.7.2, Debian 11, Python KEMI) in a
mutli-homed environment with the topos module to hide the topology.
I have noticed that when I use `set_send_socket` or `$fsn` to force the
socket it's sent from, it breaks topos and the ACK is not proxied to the
other leg of the call. I am calling record_route() *after* forcing the
socket.
If I disable TOPOS the ACK works as expected, and the signalling looks
correct.
Or, with TOPOS enabled, if I comment out set_send_socket TOPOS works as
expected.
If I disable enable_double_rr, TOPOS works as expected and the ACK is
forwarded, but without the double RR subsequent in-dialog requests use the
wrong socket.
One other observation, when I change the send socket I also see these
warnings in the logs:
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1348]: dlg_onroute():
tight matching failed for ACK with
callid='!!:MByLMlFAM.NfWxFAM.cAMxyfWjyLz.yAO.y6MxF1MxVZWG4ZMy**'/55,
ftag='2023101714101800015'/19,
ttag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36 and direction=0
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1355]: dlg_onroute():
dialog identification elements are
callid='2023101714101800015@2900-0601-0284-80'/37, caller
tag='2023101714101800015'/19, callee
tag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36
Is anyone using TOPOS + forcing the socket, or could someone advise where
I am going wrong?
Thanks in advance
Matthew
__________________________________________________________
Kamailio - Users Mailing List - Non Commercial Discussions
To unsubscribe send an email to sr-users-leave(a)lists.kamailio.org
Important: keep the mailing list in the recipients, do not reply only to
the sender!
Edit mailing list options or unsubscribe:
9:53 a.m.
Hello,
actually, there is now a mode where you can use both modules together, e.g. refer to the
docs:
https://kamailio.org/docs/modules/5.7.x/modules/topos.html#topos.p.mask_cal…
Cheers,
Henning
--
Henning Westerholt – https://skalatan.de/blog/
Kamailio services – https://gilawa.com<https://gilawa.com/>
From: Yuriy G via sr-users <sr-users(a)lists.kamailio.org>
Sent: Mittwoch, 18. Oktober 2023 09:03
To: Kamailio (SER) - Users Mailing List <sr-users(a)lists.kamailio.org>
Cc: Yuriy G <ovoshlook(a)gmail.com>
Subject: [SR-Users] Re: TOPOS + Forcing the send socket
In the header of the topic you talking about topos, but inside the messages you talking
about topoh.
They are 2 different modules. If you usr them together - they can conflictin case how they
affect message. Try use or just topoh, or just topos.
On Wed, 18 Oct 2023, 00:45 Marrold via sr-users,
<sr-users@lists.kamailio.org<mailto:sr-users@lists.kamailio.org>> wrote:
Hi all,
I've dug into this a bit more. Firstly I enabled debug logs and spotted the following
record-route header being loaded from redis:
21(28) DEBUG: PY3 {ACK}: topos_redis [topos_redis_storage.c:1079]:
tps_redis_load_dialog(): r[5]:
s[<sip:172.24.0.10:5070;sn=internal;r2=on;lr;ftag=2023101722345500002;did=c1f.18b1;nat=yes>,<sip:127.0.0.8;line=sr-N6IAzBFlWJZLWxP7WBN7z.VXN6ZQNUKJoSIBzwVLHRQ7z6fLz6g43RNQMByLMlFAM.NLMBM4W.jAMxyAMB1qCRPQ3ltEOBFZ3BFXoEt4H9IINA**>]
127.0.0.8 is the wrong IP which explains why the ACK was not being forwarded correctly. A
quick look in the source shows it's related to topoh.
I had modparam("topoh", "mask_callid", 1) in my config so I disabled
it and sure enough the ACK worked as expected which gets me a bit closer to finding the
issue.
Looking at the docs for topos and topoh it looks like things have changed since I used it
last and I should be using the following instead:
modparam("topoh", "use_mode", 1)
modparam("topos", "mask_callid", 1)
But with those configured things go really wrong:
18(25) CRITICAL: PY3 {INVITE}: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x7ffc73e3ca90 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
5(11) ERROR: <core> [core/msg_translator.c:2241]: build_req_buf_from_sip_req():
could not allocate private memory from pkg pool
5(11) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg memory for new
message
5(11) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad pointer 0x2
(out of memory block!) called from core: core/data_lump.c: free_lump(470) - ignoring
5(11) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line(): message too
short: 0 []
5(11) ERROR: <core> [core/parser/parse_fline.c:271]: parse_first_line():
parse_first_line: bad message (offset: 0)
5(11) ERROR: <core> [core/parser/msg_parser.c:748]: parse_msg(): ERROR: parse_msg:
message=<>
5(11) ERROR: <core> [core/receive.c:376]: receive_msg(): core parsing of SIP
message failed (172.24.0.21:5070/1<http://172.24.0.21:5070/1>)
3(9) ERROR: <core> [core/msg_translator.c:2241]: build_req_buf_from_sip_req():
could not allocate private memory from pkg pool
3(9) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg memory for new
message
3(9) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad pointer 0x2
(out of memory block!) called from core: core/data_lump.c: free_lump(470) - ignoring
3(9) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line(): message too
short: 0 []
Does anyone have any ideas what's going on here?
Thanks
Matthew
On Tue, Oct 17, 2023 at 6:30 PM Marrold
<kamailio@marrold.co.uk<mailto:kamailio@marrold.co.uk>> wrote:
Hi all,
I am using Kamailio (5.7.2, Debian 11, Python KEMI) in a mutli-homed environment with the
topos module to hide the topology.
I have noticed that when I use `set_send_socket` or `$fsn` to force the socket it's
sent from, it breaks topos and the ACK is not proxied to the other leg of the call. I am
calling record_route() after forcing the socket.
If I disable TOPOS the ACK works as expected, and the signalling looks correct.
Or, with TOPOS enabled, if I comment out set_send_socket TOPOS works as expected.
If I disable enable_double_rr, TOPOS works as expected and the ACK is forwarded, but
without the double RR subsequent in-dialog requests use the wrong socket.
One other observation, when I change the send socket I also see these warnings in the
logs:
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1348]: dlg_onroute(): tight matching
failed for ACK with
callid='!!:MByLMlFAM.NfWxFAM.cAMxyfWjyLz.yAO.y6MxF1MxVZWG4ZMy**'/55,
ftag='2023101714101800015'/19,
ttag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36 and direction=0
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1355]: dlg_onroute(): dialog
identification elements are callid='2023101714101800015@2900-0601-0284-80'/37,
caller tag='2023101714101800015'/19, callee
tag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36
Is anyone using TOPOS + forcing the socket, or could someone advise where I am going
wrong?
Thanks in advance
Matthew
__________________________________________________________
Kamailio - Users Mailing List - Non Commercial Discussions
To unsubscribe send an email to
sr-users-leave@lists.kamailio.org<mailto:sr-users-leave@lists.kamailio.org>
Important: keep the mailing list in the recipients, do not reply only to the sender!
Edit mailing list options or unsubscribe:
10:21 a.m.
Hi Both,
Thanks for the input. I'm now doing it the proper way:
modparam("topoh", "use_mode", 1)
modparam("topos", "mask_callid", 1)
But it's not masking the caller-id and the logs are full of errors:
18(25) CRITICAL: PY3 {INVITE}: <core> [core/mem/q_malloc.c:501]: qm_free():
BUG: bad pointer 0x7ffc73e3ca90 (out of memory block!) called from core:
core/data_lump.c: free_lump(470) - ignoring
5(11) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
5(11) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
5(11) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
5(11) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
5(11) ERROR: <core> [core/parser/parse_fline.c:271]: parse_first_line():
parse_first_line: bad message (offset: 0)
5(11) ERROR: <core> [core/parser/msg_parser.c:748]: parse_msg(): ERROR:
parse_msg: message=<>
5(11) ERROR: <core> [core/receive.c:376]: receive_msg(): core parsing of
SIP message failed (172.24.0.21:5070/1)
3(9) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
3(9) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
3(9) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
3(9) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
On Wed, 18 Oct 2023, 09:13 Henning Westerholt via sr-users, <
sr-users(a)lists.kamailio.org> wrote:
Hello,
actually, there is now a mode where you can use both modules together,
e.g. refer to the docs:
https://kamailio.org/docs/modules/5.7.x/modules/topos.html#topos.p.mask_cal…
Cheers,
Henning
--
Henning Westerholt – https://skalatan.de/blog/
Kamailio services – https://gilawa.com
*From:* Yuriy G via sr-users <sr-users(a)lists.kamailio.org>
*Sent:* Mittwoch, 18. Oktober 2023 09:03
*To:* Kamailio (SER) - Users Mailing List <sr-users(a)lists.kamailio.org>
*Cc:* Yuriy G <ovoshlook(a)gmail.com>
*Subject:* [SR-Users] Re: TOPOS + Forcing the send socket
In the header of the topic you talking about topos, but inside the
messages you talking about topoh.
They are 2 different modules. If you usr them together - they can
conflictin case how they affect message. Try use or just topoh, or just
topos.
On Wed, 18 Oct 2023, 00:45 Marrold via sr-users, <
sr-users(a)lists.kamailio.org> wrote:
Hi all,
I've dug into this a bit more. Firstly I enabled debug logs and spotted
the following record-route header being loaded from redis:
21(28) DEBUG: PY3 {ACK}: topos_redis [topos_redis_storage.c:1079]:
tps_redis_load_dialog(): r[5]: s[<
sip:172.24.0.10:5070;sn=internal;r2=on;lr;ftag=2023101722345500002;did=c1f.18b1;nat=yes
,<
sip:127.0.0.8;line=sr-N6IAzBFlWJZLWxP7WBN7z.VXN6ZQNUKJoSIBzwVLHRQ7z6fLz6g43RNQMByLMlFAM.NLMBM4W.jAMxyAMB1qCRPQ3ltEOBFZ3BFXoEt4H9IINA**
]
127.0.0.8 is the wrong IP which explains why the ACK was not being
forwarded correctly. A quick look in the source shows it's related to topoh.
I had modparam("topoh", "mask_callid", 1) in my config so I disabled
it
and sure enough the ACK worked as expected which gets me a bit closer to
finding the issue.
Looking at the docs for topos and topoh it looks like things have changed
since I used it last and I should be using the following instead:
modparam("topoh", "use_mode", 1)
modparam("topos", "mask_callid", 1)
But with those configured things go really wrong:
18(25) CRITICAL: PY3 {INVITE}: <core> [core/mem/q_malloc.c:501]:
qm_free(): BUG: bad pointer 0x7ffc73e3ca90 (out of memory block!) called
from core: core/data_lump.c: free_lump(470) - ignoring
5(11) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
5(11) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
5(11) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
5(11) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
5(11) ERROR: <core> [core/parser/parse_fline.c:271]: parse_first_line():
parse_first_line: bad message (offset: 0)
5(11) ERROR: <core> [core/parser/msg_parser.c:748]: parse_msg(): ERROR:
parse_msg: message=<>
5(11) ERROR: <core> [core/receive.c:376]: receive_msg(): core parsing of
SIP message failed (172.24.0.21:5070/1)
3(9) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
3(9) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
3(9) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
3(9) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
Does anyone have any ideas what's going on here?
Thanks
Matthew
On Tue, Oct 17, 2023 at 6:30 PM Marrold <kamailio(a)marrold.co.uk> wrote:
Hi all,
I am using Kamailio (5.7.2, Debian 11, Python KEMI) in a
mutli-homed environment with the topos module to hide the topology.
I have noticed that when I use `set_send_socket` or `$fsn` to force the
socket it's sent from, it breaks topos and the ACK is not proxied to the
other leg of the call. I am calling record_route() *after* forcing the
socket.
If I disable TOPOS the ACK works as expected, and the signalling looks
correct.
Or, with TOPOS enabled, if I comment out set_send_socket TOPOS works as
expected.
If I disable enable_double_rr, TOPOS works as expected and the ACK is
forwarded, but without the double RR subsequent in-dialog requests use the
wrong socket.
One other observation, when I change the send socket I also see these
warnings in the logs:
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1348]: dlg_onroute():
tight matching failed for ACK with
callid='!!:MByLMlFAM.NfWxFAM.cAMxyfWjyLz.yAO.y6MxF1MxVZWG4ZMy**'/55,
ftag='2023101714101800015'/19,
ttag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36 and direction=0
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1355]: dlg_onroute():
dialog identification elements are
callid='2023101714101800015@2900-0601-0284-80'/37, caller
tag='2023101714101800015'/19, callee
tag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36
Is anyone using TOPOS + forcing the socket, or could someone advise where
I am going wrong?
Thanks in advance
Matthew
__________________________________________________________
Kamailio - Users Mailing List - Non Commercial Discussions
To unsubscribe send an email to sr-users-leave(a)lists.kamailio.org
Important: keep the mailing list in the recipients, do not reply only to
the sender!
Edit mailing list options or unsubscribe:
__________________________________________________________
Kamailio - Users Mailing List - Non Commercial Discussions
To unsubscribe send an email to sr-users-leave(a)lists.kamailio.org
Important: keep the mailing list in the recipients, do not reply only to
the sender!
Edit mailing list options or unsubscribe:
11:18 a.m.
Hello,
you are running out of private memory. Please try to increase the PKG memory pool (e.g. by
changing /etc/default/kamailio or similar). You can verify with “ps aux”.
Cheers,
Henning
--
Henning Westerholt – https://skalatan.de/blog/
Kamailio services – https://gilawa.com<https://gilawa.com/>
From: Marrold via sr-users <sr-users(a)lists.kamailio.org>
Sent: Mittwoch, 18. Oktober 2023 10:21
To: Kamailio (SER) - Users Mailing List <sr-users(a)lists.kamailio.org>
Cc: Marrold <kamailio(a)marrold.co.uk>
Subject: [SR-Users] Re: TOPOS + Forcing the send socket
Hi Both,
Thanks for the input. I'm now doing it the proper way:
modparam("topoh", "use_mode", 1)
modparam("topos", "mask_callid", 1)
But it's not masking the caller-id and the logs are full of errors:
18(25) CRITICAL: PY3 {INVITE}: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x7ffc73e3ca90 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
5(11) ERROR: <core> [core/msg_translator.c:2241]: build_req_buf_from_sip_req():
could not allocate private memory from pkg pool
5(11) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg memory for new
message
5(11) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad pointer 0x2
(out of memory block!) called from core: core/data_lump.c: free_lump(470) - ignoring
5(11) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line(): message too
short: 0 []
5(11) ERROR: <core> [core/parser/parse_fline.c:271]: parse_first_line():
parse_first_line: bad message (offset: 0)
5(11) ERROR: <core> [core/parser/msg_parser.c:748]: parse_msg(): ERROR: parse_msg:
message=<>
5(11) ERROR: <core> [core/receive.c:376]: receive_msg(): core parsing of SIP
message failed (172.24.0.21:5070/1<http://172.24.0.21:5070/1>)
3(9) ERROR: <core> [core/msg_translator.c:2241]: build_req_buf_from_sip_req():
could not allocate private memory from pkg pool
3(9) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg memory for new
message
3(9) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad pointer 0x2
(out of memory block!) called from core: core/data_lump.c: free_lump(470) - ignoring
3(9) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line(): message too
short: 0 []
On Wed, 18 Oct 2023, 09:13 Henning Westerholt via sr-users,
<sr-users@lists.kamailio.org<mailto:sr-users@lists.kamailio.org>> wrote:
Hello,
actually, there is now a mode where you can use both modules together, e.g. refer to the
docs:
https://kamailio.org/docs/modules/5.7.x/modules/topos.html#topos.p.mask_cal…
Cheers,
Henning
--
Henning Westerholt – https://skalatan.de/blog/
Kamailio services – https://gilawa.com<https://gilawa.com/>
From: Yuriy G via sr-users
<sr-users@lists.kamailio.org<mailto:sr-users@lists.kamailio.org>>
Sent: Mittwoch, 18. Oktober 2023 09:03
To: Kamailio (SER) - Users Mailing List
<sr-users@lists.kamailio.org<mailto:sr-users@lists.kamailio.org>>
Cc: Yuriy G <ovoshlook@gmail.com<mailto:ovoshlook@gmail.com>>
Subject: [SR-Users] Re: TOPOS + Forcing the send socket
In the header of the topic you talking about topos, but inside the messages you talking
about topoh.
They are 2 different modules. If you usr them together - they can conflictin case how they
affect message. Try use or just topoh, or just topos.
On Wed, 18 Oct 2023, 00:45 Marrold via sr-users,
<sr-users@lists.kamailio.org<mailto:sr-users@lists.kamailio.org>> wrote:
Hi all,
I've dug into this a bit more. Firstly I enabled debug logs and spotted the following
record-route header being loaded from redis:
21(28) DEBUG: PY3 {ACK}: topos_redis [topos_redis_storage.c:1079]:
tps_redis_load_dialog(): r[5]:
s[<sip:172.24.0.10:5070;sn=internal;r2=on;lr;ftag=2023101722345500002;did=c1f.18b1;nat=yes>,<sip:127.0.0.8;line=sr-N6IAzBFlWJZLWxP7WBN7z.VXN6ZQNUKJoSIBzwVLHRQ7z6fLz6g43RNQMByLMlFAM.NLMBM4W.jAMxyAMB1qCRPQ3ltEOBFZ3BFXoEt4H9IINA**>]
127.0.0.8 is the wrong IP which explains why the ACK was not being forwarded correctly. A
quick look in the source shows it's related to topoh.
I had modparam("topoh", "mask_callid", 1) in my config so I disabled
it and sure enough the ACK worked as expected which gets me a bit closer to finding the
issue.
Looking at the docs for topos and topoh it looks like things have changed since I used it
last and I should be using the following instead:
modparam("topoh", "use_mode", 1)
modparam("topos", "mask_callid", 1)
But with those configured things go really wrong:
18(25) CRITICAL: PY3 {INVITE}: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x7ffc73e3ca90 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
5(11) ERROR: <core> [core/msg_translator.c:2241]: build_req_buf_from_sip_req():
could not allocate private memory from pkg pool
5(11) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg memory for new
message
5(11) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad pointer 0x2
(out of memory block!) called from core: core/data_lump.c: free_lump(470) - ignoring
5(11) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line(): message too
short: 0 []
5(11) ERROR: <core> [core/parser/parse_fline.c:271]: parse_first_line():
parse_first_line: bad message (offset: 0)
5(11) ERROR: <core> [core/parser/msg_parser.c:748]: parse_msg(): ERROR: parse_msg:
message=<>
5(11) ERROR: <core> [core/receive.c:376]: receive_msg(): core parsing of SIP
message failed (172.24.0.21:5070/1<http://172.24.0.21:5070/1>)
3(9) ERROR: <core> [core/msg_translator.c:2241]: build_req_buf_from_sip_req():
could not allocate private memory from pkg pool
3(9) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg memory for new
message
3(9) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad pointer 0x2
(out of memory block!) called from core: core/data_lump.c: free_lump(470) - ignoring
3(9) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line(): message too
short: 0 []
Does anyone have any ideas what's going on here?
Thanks
Matthew
On Tue, Oct 17, 2023 at 6:30 PM Marrold
<kamailio@marrold.co.uk<mailto:kamailio@marrold.co.uk>> wrote:
Hi all,
I am using Kamailio (5.7.2, Debian 11, Python KEMI) in a mutli-homed environment with the
topos module to hide the topology.
I have noticed that when I use `set_send_socket` or `$fsn` to force the socket it's
sent from, it breaks topos and the ACK is not proxied to the other leg of the call. I am
calling record_route() after forcing the socket.
If I disable TOPOS the ACK works as expected, and the signalling looks correct.
Or, with TOPOS enabled, if I comment out set_send_socket TOPOS works as expected.
If I disable enable_double_rr, TOPOS works as expected and the ACK is forwarded, but
without the double RR subsequent in-dialog requests use the wrong socket.
One other observation, when I change the send socket I also see these warnings in the
logs:
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1348]: dlg_onroute(): tight matching
failed for ACK with
callid='!!:MByLMlFAM.NfWxFAM.cAMxyfWjyLz.yAO.y6MxF1MxVZWG4ZMy**'/55,
ftag='2023101714101800015'/19,
ttag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36 and direction=0
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1355]: dlg_onroute(): dialog
identification elements are callid='2023101714101800015@2900-0601-0284-80'/37,
caller tag='2023101714101800015'/19, callee
tag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36
Is anyone using TOPOS + forcing the socket, or could someone advise where I am going
wrong?
Thanks in advance
Matthew
__________________________________________________________
Kamailio - Users Mailing List - Non Commercial Discussions
To unsubscribe send an email to
sr-users-leave@lists.kamailio.org<mailto:sr-users-leave@lists.kamailio.org>
Important: keep the mailing list in the recipients, do not reply only to the sender!
Edit mailing list options or unsubscribe:
__________________________________________________________
Kamailio - Users Mailing List - Non Commercial Discussions
To unsubscribe send an email to
sr-users-leave@lists.kamailio.org<mailto:sr-users-leave@lists.kamailio.org>
Important: keep the mailing list in the recipients, do not reply only to the sender!
Edit mailing list options or unsubscribe:
19 Oct
19 Oct
2:05 a.m.
Hi Henning,
I bumped the PKG memory up to 24MB all the way up to 128MB and I still get
the same issue.
Thanks
Matthew
On Wed, Oct 18, 2023 at 10:18 AM Henning Westerholt <hw(a)gilawa.com> wrote:
Hello,
you are running out of private memory. Please try to increase the PKG
memory pool (e.g. by changing /etc/default/kamailio or similar). You can
verify with “ps aux”.
Cheers,
Henning
--
Henning Westerholt – https://skalatan.de/blog/
Kamailio services – https://gilawa.com
*From:* Marrold via sr-users <sr-users(a)lists.kamailio.org>
*Sent:* Mittwoch, 18. Oktober 2023 10:21
*To:* Kamailio (SER) - Users Mailing List <sr-users(a)lists.kamailio.org>
*Cc:* Marrold <kamailio(a)marrold.co.uk>
*Subject:* [SR-Users] Re: TOPOS + Forcing the send socket
Hi Both,
Thanks for the input. I'm now doing it the proper way:
modparam("topoh", "use_mode", 1)
modparam("topos", "mask_callid", 1)
But it's not masking the caller-id and the logs are full of errors:
18(25) CRITICAL: PY3 {INVITE}: <core> [core/mem/q_malloc.c:501]:
qm_free(): BUG: bad pointer 0x7ffc73e3ca90 (out of memory block!) called
from core: core/data_lump.c: free_lump(470) - ignoring
5(11) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
5(11) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
5(11) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
5(11) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
5(11) ERROR: <core> [core/parser/parse_fline.c:271]: parse_first_line():
parse_first_line: bad message (offset: 0)
5(11) ERROR: <core> [core/parser/msg_parser.c:748]: parse_msg(): ERROR:
parse_msg: message=<>
5(11) ERROR: <core> [core/receive.c:376]: receive_msg(): core parsing of
SIP message failed (172.24.0.21:5070/1)
3(9) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
3(9) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
3(9) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
3(9) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
On Wed, 18 Oct 2023, 09:13 Henning Westerholt via sr-users, <
sr-users(a)lists.kamailio.org> wrote:
Hello,
actually, there is now a mode where you can use both modules together,
e.g. refer to the docs:
https://kamailio.org/docs/modules/5.7.x/modules/topos.html#topos.p.mask_cal…
Cheers,
Henning
--
Henning Westerholt – https://skalatan.de/blog/
Kamailio services – https://gilawa.com
*From:* Yuriy G via sr-users <sr-users(a)lists.kamailio.org>
*Sent:* Mittwoch, 18. Oktober 2023 09:03
*To:* Kamailio (SER) - Users Mailing List <sr-users(a)lists.kamailio.org>
*Cc:* Yuriy G <ovoshlook(a)gmail.com>
*Subject:* [SR-Users] Re: TOPOS + Forcing the send socket
In the header of the topic you talking about topos, but inside the
messages you talking about topoh.
They are 2 different modules. If you usr them together - they can
conflictin case how they affect message. Try use or just topoh, or just
topos.
On Wed, 18 Oct 2023, 00:45 Marrold via sr-users, <
sr-users(a)lists.kamailio.org> wrote:
Hi all,
I've dug into this a bit more. Firstly I enabled debug logs and spotted
the following record-route header being loaded from redis:
21(28) DEBUG: PY3 {ACK}: topos_redis [topos_redis_storage.c:1079]:
tps_redis_load_dialog(): r[5]: s[<
sip:172.24.0.10:5070;sn=internal;r2=on;lr;ftag=2023101722345500002;did=c1f.18b1;nat=yes
,<
sip:127.0.0.8;line=sr-N6IAzBFlWJZLWxP7WBN7z.VXN6ZQNUKJoSIBzwVLHRQ7z6fLz6g43RNQMByLMlFAM.NLMBM4W.jAMxyAMB1qCRPQ3ltEOBFZ3BFXoEt4H9IINA**
]
127.0.0.8 is the wrong IP which explains why the ACK was not being
forwarded correctly. A quick look in the source shows it's related to topoh.
I had modparam("topoh", "mask_callid", 1) in my config so I disabled
it
and sure enough the ACK worked as expected which gets me a bit closer to
finding the issue.
Looking at the docs for topos and topoh it looks like things have changed
since I used it last and I should be using the following instead:
modparam("topoh", "use_mode", 1)
modparam("topos", "mask_callid", 1)
But with those configured things go really wrong:
18(25) CRITICAL: PY3 {INVITE}: <core> [core/mem/q_malloc.c:501]:
qm_free(): BUG: bad pointer 0x7ffc73e3ca90 (out of memory block!) called
from core: core/data_lump.c: free_lump(470) - ignoring
5(11) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
5(11) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
5(11) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
5(11) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
5(11) ERROR: <core> [core/parser/parse_fline.c:271]: parse_first_line():
parse_first_line: bad message (offset: 0)
5(11) ERROR: <core> [core/parser/msg_parser.c:748]: parse_msg(): ERROR:
parse_msg: message=<>
5(11) ERROR: <core> [core/receive.c:376]: receive_msg(): core parsing of
SIP message failed (172.24.0.21:5070/1)
3(9) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
3(9) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
3(9) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
3(9) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
Does anyone have any ideas what's going on here?
Thanks
Matthew
On Tue, Oct 17, 2023 at 6:30 PM Marrold <kamailio(a)marrold.co.uk> wrote:
Hi all,
I am using Kamailio (5.7.2, Debian 11, Python KEMI) in a
mutli-homed environment with the topos module to hide the topology.
I have noticed that when I use `set_send_socket` or `$fsn` to force the
socket it's sent from, it breaks topos and the ACK is not proxied to the
other leg of the call. I am calling record_route() *after* forcing the
socket.
If I disable TOPOS the ACK works as expected, and the signalling looks
correct.
Or, with TOPOS enabled, if I comment out set_send_socket TOPOS works as
expected.
If I disable enable_double_rr, TOPOS works as expected and the ACK is
forwarded, but without the double RR subsequent in-dialog requests use the
wrong socket.
One other observation, when I change the send socket I also see these
warnings in the logs:
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1348]: dlg_onroute():
tight matching failed for ACK with
callid='!!:MByLMlFAM.NfWxFAM.cAMxyfWjyLz.yAO.y6MxF1MxVZWG4ZMy**'/55,
ftag='2023101714101800015'/19,
ttag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36 and direction=0
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1355]: dlg_onroute():
dialog identification elements are
callid='2023101714101800015@2900-0601-0284-80'/37, caller
tag='2023101714101800015'/19, callee
tag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36
Is anyone using TOPOS + forcing the socket, or could someone advise where
I am going wrong?
Thanks in advance
Matthew
__________________________________________________________
Kamailio - Users Mailing List - Non Commercial Discussions
To unsubscribe send an email to sr-users-leave(a)lists.kamailio.org
Important: keep the mailing list in the recipients, do not reply only to
the sender!
Edit mailing list options or unsubscribe:
__________________________________________________________
Kamailio - Users Mailing List - Non Commercial Discussions
To unsubscribe send an email to sr-users-leave(a)lists.kamailio.org
Important: keep the mailing list in the recipients, do not reply only to
the sender!
Edit mailing list options or unsubscribe:
8:37 a.m.
Hello,
this is quite too much, then you are probably having a memory leak. Have a look e.g. to
this for debugging help:
https://www.kamailio.org/wiki/tutorials/troubleshooting/memory
Cheers,
Henning
--
Henning Westerholt – https://skalatan.de/blog/
Kamailio services – https://gilawa.com<https://gilawa.com/>
From: Marrold <kamailio(a)marrold.co.uk>
Sent: Donnerstag, 19. Oktober 2023 02:05
To: Henning Westerholt <hw(a)gilawa.com>
Cc: Kamailio (SER) - Users Mailing List <sr-users(a)lists.kamailio.org>
Subject: Re: [SR-Users] Re: TOPOS + Forcing the send socket
Hi Henning,
I bumped the PKG memory up to 24MB all the way up to 128MB and I still get the same
issue.
Thanks
Matthew
On Wed, Oct 18, 2023 at 10:18 AM Henning Westerholt
<hw@gilawa.com<mailto:hw@gilawa.com>> wrote:
Hello,
you are running out of private memory. Please try to increase the PKG memory pool (e.g. by
changing /etc/default/kamailio or similar). You can verify with “ps aux”.
Cheers,
Henning
--
Henning Westerholt – https://skalatan.de/blog/
Kamailio services – https://gilawa.com<https://gilawa.com/>
From: Marrold via sr-users
<sr-users@lists.kamailio.org<mailto:sr-users@lists.kamailio.org>>
Sent: Mittwoch, 18. Oktober 2023 10:21
To: Kamailio (SER) - Users Mailing List
<sr-users@lists.kamailio.org<mailto:sr-users@lists.kamailio.org>>
Cc: Marrold <kamailio@marrold.co.uk<mailto:kamailio@marrold.co.uk>>
Subject: [SR-Users] Re: TOPOS + Forcing the send socket
Hi Both,
Thanks for the input. I'm now doing it the proper way:
modparam("topoh", "use_mode", 1)
modparam("topos", "mask_callid", 1)
But it's not masking the caller-id and the logs are full of errors:
18(25) CRITICAL: PY3 {INVITE}: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x7ffc73e3ca90 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
5(11) ERROR: <core> [core/msg_translator.c:2241]: build_req_buf_from_sip_req():
could not allocate private memory from pkg pool
5(11) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg memory for new
message
5(11) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad pointer 0x2
(out of memory block!) called from core: core/data_lump.c: free_lump(470) - ignoring
5(11) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line(): message too
short: 0 []
5(11) ERROR: <core> [core/parser/parse_fline.c:271]: parse_first_line():
parse_first_line: bad message (offset: 0)
5(11) ERROR: <core> [core/parser/msg_parser.c:748]: parse_msg(): ERROR: parse_msg:
message=<>
5(11) ERROR: <core> [core/receive.c:376]: receive_msg(): core parsing of SIP
message failed (172.24.0.21:5070/1<http://172.24.0.21:5070/1>)
3(9) ERROR: <core> [core/msg_translator.c:2241]: build_req_buf_from_sip_req():
could not allocate private memory from pkg pool
3(9) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg memory for new
message
3(9) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad pointer 0x2
(out of memory block!) called from core: core/data_lump.c: free_lump(470) - ignoring
3(9) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line(): message too
short: 0 []
On Wed, 18 Oct 2023, 09:13 Henning Westerholt via sr-users,
<sr-users@lists.kamailio.org<mailto:sr-users@lists.kamailio.org>> wrote:
Hello,
actually, there is now a mode where you can use both modules together, e.g. refer to the
docs:
https://kamailio.org/docs/modules/5.7.x/modules/topos.html#topos.p.mask_cal…
Cheers,
Henning
--
Henning Westerholt – https://skalatan.de/blog/
Kamailio services – https://gilawa.com<https://gilawa.com/>
From: Yuriy G via sr-users
<sr-users@lists.kamailio.org<mailto:sr-users@lists.kamailio.org>>
Sent: Mittwoch, 18. Oktober 2023 09:03
To: Kamailio (SER) - Users Mailing List
<sr-users@lists.kamailio.org<mailto:sr-users@lists.kamailio.org>>
Cc: Yuriy G <ovoshlook@gmail.com<mailto:ovoshlook@gmail.com>>
Subject: [SR-Users] Re: TOPOS + Forcing the send socket
In the header of the topic you talking about topos, but inside the messages you talking
about topoh.
They are 2 different modules. If you usr them together - they can conflictin case how they
affect message. Try use or just topoh, or just topos.
On Wed, 18 Oct 2023, 00:45 Marrold via sr-users,
<sr-users@lists.kamailio.org<mailto:sr-users@lists.kamailio.org>> wrote:
Hi all,
I've dug into this a bit more. Firstly I enabled debug logs and spotted the following
record-route header being loaded from redis:
21(28) DEBUG: PY3 {ACK}: topos_redis [topos_redis_storage.c:1079]:
tps_redis_load_dialog(): r[5]:
s[<sip:172.24.0.10:5070;sn=internal;r2=on;lr;ftag=2023101722345500002;did=c1f.18b1;nat=yes>,<sip:127.0.0.8;line=sr-N6IAzBFlWJZLWxP7WBN7z.VXN6ZQNUKJoSIBzwVLHRQ7z6fLz6g43RNQMByLMlFAM.NLMBM4W.jAMxyAMB1qCRPQ3ltEOBFZ3BFXoEt4H9IINA**>]
127.0.0.8 is the wrong IP which explains why the ACK was not being forwarded correctly. A
quick look in the source shows it's related to topoh.
I had modparam("topoh", "mask_callid", 1) in my config so I disabled
it and sure enough the ACK worked as expected which gets me a bit closer to finding the
issue.
Looking at the docs for topos and topoh it looks like things have changed since I used it
last and I should be using the following instead:
modparam("topoh", "use_mode", 1)
modparam("topos", "mask_callid", 1)
But with those configured things go really wrong:
18(25) CRITICAL: PY3 {INVITE}: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x7ffc73e3ca90 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
5(11) ERROR: <core> [core/msg_translator.c:2241]: build_req_buf_from_sip_req():
could not allocate private memory from pkg pool
5(11) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg memory for new
message
5(11) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad pointer 0x2
(out of memory block!) called from core: core/data_lump.c: free_lump(470) - ignoring
5(11) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line(): message too
short: 0 []
5(11) ERROR: <core> [core/parser/parse_fline.c:271]: parse_first_line():
parse_first_line: bad message (offset: 0)
5(11) ERROR: <core> [core/parser/msg_parser.c:748]: parse_msg(): ERROR: parse_msg:
message=<>
5(11) ERROR: <core> [core/receive.c:376]: receive_msg(): core parsing of SIP
message failed (172.24.0.21:5070/1<http://172.24.0.21:5070/1>)
3(9) ERROR: <core> [core/msg_translator.c:2241]: build_req_buf_from_sip_req():
could not allocate private memory from pkg pool
3(9) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg memory for new
message
3(9) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad pointer 0x2
(out of memory block!) called from core: core/data_lump.c: free_lump(470) - ignoring
3(9) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line(): message too
short: 0 []
Does anyone have any ideas what's going on here?
Thanks
Matthew
On Tue, Oct 17, 2023 at 6:30 PM Marrold
<kamailio@marrold.co.uk<mailto:kamailio@marrold.co.uk>> wrote:
Hi all,
I am using Kamailio (5.7.2, Debian 11, Python KEMI) in a mutli-homed environment with the
topos module to hide the topology.
I have noticed that when I use `set_send_socket` or `$fsn` to force the socket it's
sent from, it breaks topos and the ACK is not proxied to the other leg of the call. I am
calling record_route() after forcing the socket.
If I disable TOPOS the ACK works as expected, and the signalling looks correct.
Or, with TOPOS enabled, if I comment out set_send_socket TOPOS works as expected.
If I disable enable_double_rr, TOPOS works as expected and the ACK is forwarded, but
without the double RR subsequent in-dialog requests use the wrong socket.
One other observation, when I change the send socket I also see these warnings in the
logs:
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1348]: dlg_onroute(): tight matching
failed for ACK with
callid='!!:MByLMlFAM.NfWxFAM.cAMxyfWjyLz.yAO.y6MxF1MxVZWG4ZMy**'/55,
ftag='2023101714101800015'/19,
ttag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36 and direction=0
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1355]: dlg_onroute(): dialog
identification elements are callid='2023101714101800015@2900-0601-0284-80'/37,
caller tag='2023101714101800015'/19, callee
tag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36
Is anyone using TOPOS + forcing the socket, or could someone advise where I am going
wrong?
Thanks in advance
Matthew
__________________________________________________________
Kamailio - Users Mailing List - Non Commercial Discussions
To unsubscribe send an email to
sr-users-leave@lists.kamailio.org<mailto:sr-users-leave@lists.kamailio.org>
Important: keep the mailing list in the recipients, do not reply only to the sender!
Edit mailing list options or unsubscribe:
__________________________________________________________
Kamailio - Users Mailing List - Non Commercial Discussions
To unsubscribe send an email to
sr-users-leave@lists.kamailio.org<mailto:sr-users-leave@lists.kamailio.org>
Important: keep the mailing list in the recipients, do not reply only to the sender!
Edit mailing list options or unsubscribe:
10:22 a.m.
Thanks Henning I'll give that a try.
For additional context this happens when a single call traverses the
system, fresh after a restart, so if there's a memory leak it seems to
happen rapidly whilst processing a single SIP message
Cheers
Matthew
On Thu, 19 Oct 2023, 07:37 Henning Westerholt, <hw(a)gilawa.com> wrote:
Hello,
this is quite too much, then you are probably having a memory leak. Have a
look e.g. to this for debugging help:
https://www.kamailio.org/wiki/tutorials/troubleshooting/memory
Cheers,
Henning
--
Henning Westerholt – https://skalatan.de/blog/
Kamailio services – https://gilawa.com
*From:* Marrold <kamailio(a)marrold.co.uk>
*Sent:* Donnerstag, 19. Oktober 2023 02:05
*To:* Henning Westerholt <hw(a)gilawa.com>
*Cc:* Kamailio (SER) - Users Mailing List <sr-users(a)lists.kamailio.org>
*Subject:* Re: [SR-Users] Re: TOPOS + Forcing the send socket
Hi Henning,
I bumped the PKG memory up to 24MB all the way up to 128MB and I still get
the same issue.
Thanks
Matthew
On Wed, Oct 18, 2023 at 10:18 AM Henning Westerholt <hw(a)gilawa.com> wrote:
Hello,
you are running out of private memory. Please try to increase the PKG
memory pool (e.g. by changing /etc/default/kamailio or similar). You can
verify with “ps aux”.
Cheers,
Henning
--
Henning Westerholt – https://skalatan.de/blog/
Kamailio services – https://gilawa.com
*From:* Marrold via sr-users <sr-users(a)lists.kamailio.org>
*Sent:* Mittwoch, 18. Oktober 2023 10:21
*To:* Kamailio (SER) - Users Mailing List <sr-users(a)lists.kamailio.org>
*Cc:* Marrold <kamailio(a)marrold.co.uk>
*Subject:* [SR-Users] Re: TOPOS + Forcing the send socket
Hi Both,
Thanks for the input. I'm now doing it the proper way:
modparam("topoh", "use_mode", 1)
modparam("topos", "mask_callid", 1)
But it's not masking the caller-id and the logs are full of errors:
18(25) CRITICAL: PY3 {INVITE}: <core> [core/mem/q_malloc.c:501]:
qm_free(): BUG: bad pointer 0x7ffc73e3ca90 (out of memory block!) called
from core: core/data_lump.c: free_lump(470) - ignoring
5(11) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
5(11) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
5(11) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
5(11) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
5(11) ERROR: <core> [core/parser/parse_fline.c:271]: parse_first_line():
parse_first_line: bad message (offset: 0)
5(11) ERROR: <core> [core/parser/msg_parser.c:748]: parse_msg(): ERROR:
parse_msg: message=<>
5(11) ERROR: <core> [core/receive.c:376]: receive_msg(): core parsing of
SIP message failed (172.24.0.21:5070/1)
3(9) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
3(9) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
3(9) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
3(9) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
On Wed, 18 Oct 2023, 09:13 Henning Westerholt via sr-users, <
sr-users(a)lists.kamailio.org> wrote:
Hello,
actually, there is now a mode where you can use both modules together,
e.g. refer to the docs:
https://kamailio.org/docs/modules/5.7.x/modules/topos.html#topos.p.mask_cal…
Cheers,
Henning
--
Henning Westerholt – https://skalatan.de/blog/
Kamailio services – https://gilawa.com
*From:* Yuriy G via sr-users <sr-users(a)lists.kamailio.org>
*Sent:* Mittwoch, 18. Oktober 2023 09:03
*To:* Kamailio (SER) - Users Mailing List <sr-users(a)lists.kamailio.org>
*Cc:* Yuriy G <ovoshlook(a)gmail.com>
*Subject:* [SR-Users] Re: TOPOS + Forcing the send socket
In the header of the topic you talking about topos, but inside the
messages you talking about topoh.
They are 2 different modules. If you usr them together - they can
conflictin case how they affect message. Try use or just topoh, or just
topos.
On Wed, 18 Oct 2023, 00:45 Marrold via sr-users, <
sr-users(a)lists.kamailio.org> wrote:
Hi all,
I've dug into this a bit more. Firstly I enabled debug logs and spotted
the following record-route header being loaded from redis:
21(28) DEBUG: PY3 {ACK}: topos_redis [topos_redis_storage.c:1079]:
tps_redis_load_dialog(): r[5]: s[<
sip:172.24.0.10:5070;sn=internal;r2=on;lr;ftag=2023101722345500002;did=c1f.18b1;nat=yes
,<
sip:127.0.0.8;line=sr-N6IAzBFlWJZLWxP7WBN7z.VXN6ZQNUKJoSIBzwVLHRQ7z6fLz6g43RNQMByLMlFAM.NLMBM4W.jAMxyAMB1qCRPQ3ltEOBFZ3BFXoEt4H9IINA**
]
127.0.0.8 is the wrong IP which explains why the ACK was not being
forwarded correctly. A quick look in the source shows it's related to topoh.
I had modparam("topoh", "mask_callid", 1) in my config so I disabled
it
and sure enough the ACK worked as expected which gets me a bit closer to
finding the issue.
Looking at the docs for topos and topoh it looks like things have changed
since I used it last and I should be using the following instead:
modparam("topoh", "use_mode", 1)
modparam("topos", "mask_callid", 1)
But with those configured things go really wrong:
18(25) CRITICAL: PY3 {INVITE}: <core> [core/mem/q_malloc.c:501]:
qm_free(): BUG: bad pointer 0x7ffc73e3ca90 (out of memory block!) called
from core: core/data_lump.c: free_lump(470) - ignoring
5(11) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
5(11) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
5(11) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
5(11) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
5(11) ERROR: <core> [core/parser/parse_fline.c:271]: parse_first_line():
parse_first_line: bad message (offset: 0)
5(11) ERROR: <core> [core/parser/msg_parser.c:748]: parse_msg(): ERROR:
parse_msg: message=<>
5(11) ERROR: <core> [core/receive.c:376]: receive_msg(): core parsing of
SIP message failed (172.24.0.21:5070/1)
3(9) ERROR: <core> [core/msg_translator.c:2241]:
build_req_buf_from_sip_req(): could not allocate private memory from pkg
pool
3(9) ERROR: topos [topos_mod.c:518]: tps_msg_received(): not enough pkg
memory for new message
3(9) CRITICAL: <core> [core/mem/q_malloc.c:501]: qm_free(): BUG: bad
pointer 0x2 (out of memory block!) called from core: core/data_lump.c:
free_lump(470) - ignoring
3(9) INFO: <core> [core/parser/parse_fline.c:80]: parse_first_line():
message too short: 0 []
Does anyone have any ideas what's going on here?
Thanks
Matthew
On Tue, Oct 17, 2023 at 6:30 PM Marrold <kamailio(a)marrold.co.uk> wrote:
Hi all,
I am using Kamailio (5.7.2, Debian 11, Python KEMI) in a
mutli-homed environment with the topos module to hide the topology.
I have noticed that when I use `set_send_socket` or `$fsn` to force the
socket it's sent from, it breaks topos and the ACK is not proxied to the
other leg of the call. I am calling record_route() *after* forcing the
socket.
If I disable TOPOS the ACK works as expected, and the signalling looks
correct.
Or, with TOPOS enabled, if I comment out set_send_socket TOPOS works as
expected.
If I disable enable_double_rr, TOPOS works as expected and the ACK is
forwarded, but without the double RR subsequent in-dialog requests use the
wrong socket.
One other observation, when I change the send socket I also see these
warnings in the logs:
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1348]: dlg_onroute():
tight matching failed for ACK with
callid='!!:MByLMlFAM.NfWxFAM.cAMxyfWjyLz.yAO.y6MxF1MxVZWG4ZMy**'/55,
ftag='2023101714101800015'/19,
ttag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36 and direction=0
24(32) WARNING: PY3 {ACK}: dialog [dlg_handlers.c:1355]: dlg_onroute():
dialog identification elements are
callid='2023101714101800015@2900-0601-0284-80'/37, caller
tag='2023101714101800015'/19, callee
tag='2f55349a-2c59-4e37-bf58-fd84fb69ece9'/36
Is anyone using TOPOS + forcing the socket, or could someone advise where
I am going wrong?
Thanks in advance
Matthew
__________________________________________________________
Kamailio - Users Mailing List - Non Commercial Discussions
To unsubscribe send an email to sr-users-leave(a)lists.kamailio.org
Important: keep the mailing list in the recipients, do not reply only to
the sender!
Edit mailing list options or unsubscribe:
__________________________________________________________
Kamailio - Users Mailing List - Non Commercial Discussions
To unsubscribe send an email to sr-users-leave(a)lists.kamailio.org
Important: keep the mailing list in the recipients, do not reply only to
the sender!
Edit mailing list options or unsubscribe:
214
days inactive
216
days old
8 comments
3 participants
participants (3)
-
Henning Westerholt -
Marrold -
Yuriy G