My SIP proxy didn't start with Kamailio 5.8 using the same config that starts OK with 5.7. The error is:
2024-03-11T06:56:06.457107+02:00 lohi /usr/bin/sip-proxy[2358954]: DEBUG: acc [acc_mod.c:362]: parse_failed_filter(): failed_filter 0 = 407 2024-03-11T06:56:06.457265+02:00 lohi /usr/bin/sip-proxy[2358954]: ERROR: acc [acc_mod.c:369]: parse_failed_filter(): response code is not followed by comma or end of string 2024-03-11T06:56:06.457436+02:00 lohi /usr/bin/sip-proxy[2358954]: ERROR: acc [acc_mod.c:439]: mod_init(): failed to parse failed_filter param
The relevant params are:
modparam("acc", "failed_transaction_flag", 8) modparam("acc", "failed_filter", "407")
I didn't find any difference in acc_mod.c source code between 5.7 and 5.8 and parse_failed_filter function looked OK.
Then I added a debug statement in the beginning of parse_failed_filter function:
LM_DBG("parsing failed_filter %s\n", s);
and got to syslog:
cc_mod.c:341]: parse_failed_filter(): parsing failed_filter 407rmissions|pua|rtpengine)$
That explains the parse_failed_filter error, but where does that bogus param value come from?
Before acc params I have:
modparam("auth_db|dialplan|domain|htable|lcr|msilo|mtree|permissions|pua|rtpengine", "db_url", "mysql://xxxx/sip_proxy") modparam("registrar|nathelper", "received_avp", "$avp(received_uri)")
So it from there, but why? Is there a bug somewhere or is my SIP proxy running out of memory or something? Before the above error messages, there are no other error messages.
-- Juha
Hello,
what are the operating system and the compiler used?
Cheers, Daniel
On 11.03.24 07:45, Juha Heinanen via sr-users wrote:
My SIP proxy didn't start with Kamailio 5.8 using the same config that starts OK with 5.7. The error is:
2024-03-11T06:56:06.457107+02:00 lohi /usr/bin/sip-proxy[2358954]: DEBUG: acc [acc_mod.c:362]: parse_failed_filter(): failed_filter 0 = 407 2024-03-11T06:56:06.457265+02:00 lohi /usr/bin/sip-proxy[2358954]: ERROR: acc [acc_mod.c:369]: parse_failed_filter(): response code is not followed by comma or end of string 2024-03-11T06:56:06.457436+02:00 lohi /usr/bin/sip-proxy[2358954]: ERROR: acc [acc_mod.c:439]: mod_init(): failed to parse failed_filter param
The relevant params are:
modparam("acc", "failed_transaction_flag", 8) modparam("acc", "failed_filter", "407")
I didn't find any difference in acc_mod.c source code between 5.7 and 5.8 and parse_failed_filter function looked OK.
Then I added a debug statement in the beginning of parse_failed_filter function:
LM_DBG("parsing failed_filter %s\n", s);
and got to syslog:
cc_mod.c:341]: parse_failed_filter(): parsing failed_filter 407rmissions|pua|rtpengine)$
That explains the parse_failed_filter error, but where does that bogus param value come from?
Before acc params I have:
modparam("auth_db|dialplan|domain|htable|lcr|msilo|mtree|permissions|pua|rtpengine", "db_url", "mysql://xxxx/sip_proxy") modparam("registrar|nathelper", "received_avp", "$avp(received_uri)")
So it from there, but why? Is there a bug somewhere or is my SIP proxy running out of memory or something? Before the above error messages, there are no other error messages.
-- Juha
Kamailio - Users Mailing List - Non Commercial Discussions To unsubscribe send an email to 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:
On 11.03.24 20:45, Juha Heinanen via sr-users wrote:
Daniel-Constantin Mierla writes:
what are the operating system and the compiler used?
Sorry, Debian 12 using gcc version 12.2.0.
It might be a regression by commit 675f6c02248, can you try with latest master or with the patch of the commit:
- https://github.com/kamailio/kamailio/commit/4dc37aee853bbcf9d7fdb96245b76aaa...
If it works, then you can go ahead and backport to 5.8 branch.
Cheers, Daniel
Daniel-Constantin Mierla writes:
It might be a regression by commit 675f6c02248, can you try with latest master or with the patch of the commit:
- https://github.com/kamailio/kamailio/commit/4dc37aee853bbcf9d7fdb96245b76aaa...
I tried with latest master and it started OK. Pretty serious stuff. Perhaps a new tag 5.8.1 is needed.
-- Juha
On 12.03.24 03:13, Juha Heinanen wrote:
Daniel-Constantin Mierla writes:
It might be a regression by commit 675f6c02248, can you try with latest master or with the patch of the commit:
- https://github.com/kamailio/kamailio/commit/4dc37aee853bbcf9d7fdb96245b76aaa...
I tried with latest master and it started OK. Pretty serious stuff. Perhaps a new tag 5.8.1 is needed.
It will be a 5.8.1 soon, but we have to wait and see what else it pops up, usually after a .0 release, a bunch of things need tuning, releasing after each commit is not feasible. We also have nightly builds for debian/ubuntu and source installation should use the branch 5.8, not tags.
On the other hand, it might not be as bad as it seems, it is only for startup and only when the a previous modparam (or some other operation) frees a chunk that can be reused by next one. Also, it affects only PARAM_STRING, not PARAM_STR. That's the reason it was not discovered by testing with the default config file or other configs (at least I run master in a testing environment with many more modules loaded than default kamailio.cfg) -- the regression is rather old, from June 2023. The main docker-based tests were also ok.
Cheers, Daniel
-
Daniel-Constantin Mierla -
Juha Heinanen