On 10/29/10 5:40 PM, Sergey Okhapkin wrote:
I can't test now - high volume traffic is running. But I looked at the diff, it seems to me OK, I see no reason why it could not work:-)
Because of the law, the Murphy law ;-)
Probably I will test sometime next week if you don't do it before and then backport.
Cheers, Daniel
On Friday 29 October 2010, you wrote:
On 10/29/10 4:33 PM, Daniel-Constantin Mierla wrote:
On 10/29/10 4:09 PM, Sergey Okhapkin wrote:
I just realized that 3.1 executes main routing block when broken SIP message is received! But sanity_check() function returns 0 on failed checks and script execution stops, my favorite xlog() is not executed.
OK, that can be fixed very easy. Iirc, in some cases you don't get the sip message in the config if it is really broken, but I am not sure if error_route was executed for that case anyhow.
I committed on git master, id:
39a8108c62a2beafba18417613b9660a8abef86a
Can you take the patch and test it on 3.1? If all is fine, then I can backport - even the default cfg expects non-drop behavior since it attempts to print a log message there.
Cheers, Daniel
Cheers, Daniel
Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: ERROR:<core> [parser/msg_parser.c:177]: ERROR: get_hdr_field: bad to header Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: INFO:<core> [parser/msg_parser.c:351]: ERROR: bad header field [To: Chi Nhung<sip: ] Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: ERROR:<core> [parser/msg_parser.c:177]: ERROR: get_hdr_field: bad to header Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: INFO:<core> [parser/msg_parser.c:351]: ERROR: bad header field [To: Chi Nhung<sip: ] Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: ERROR:<core> [parser/msg_parser.c:177]: ERROR: get_hdr_field: bad to header Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: INFO:<core> [parser/msg_parser.c:351]: ERROR: bad header field [To: Chi Nhung<sip: ] Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: ERROR:<core> [msg_translator.c:1946]: ERROR: build_res_buf_from_sip_req: alas, parse _headers failed Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: WARNING: sanity [sanity.c:253]: sanity_check(): check_required_headers(): failed to s end 400 via sl reply
On Friday 29 October 2010, Daniel-Constantin Mierla wrote:
On 10/29/10 11:53 AM, Sergey Okhapkin wrote:
Daniel, I think this could be implemented without re-introduction of error_route,
it will not be re-introduced for sure, we have now a different core framework which care reuse event_route mechanism.
In the past the error route was executed when the error was discovered, so it several places of the code, and I am not sure that the core can continue up to config execution of something is really broken.
I will have it in mind when I get some spare time.
Cheers, Daniel
what about core parameter route_on_error=yes/no (default no)? Ifroute on error is enabled, then sanity module will do the trick:
route{ if(!sanity_check("1159")) { xlog("L_INFO","Bad message from $proto:$si:$sp\n$mb\n"); break; } ....
It will be up to script writer what to do in main routing block if message parsing fails.
On Friday 29 October 2010, Sergey Okhapkin wrote: > It would be nice to get error_route back. The route was trivial in my > cfg file: > > error_route { > xlog("L_INFO","Bad message from $proto:$si:$sp\n$mb\n"); > drop(); > } > > On Friday 29 October 2010, Daniel-Constantin Mierla wrote: >> On 10/28/10 1:41 PM, Sergey Okhapkin wrote: >>> I've found one more feature missed in 3.1 - error_route is >>> eliminated >>> and sanity module added. But if received SIP message is >>> malformed and >>> can't be parsed, routing script is not executed and I have no >>> way to >>> log the message to kamailio log. I did it in error_route before. >>> >>> This logging is important to me to support customers. SIP ALG in >>> routers often breaks Via line :-( >> Since I developed error_route, I remember that it handled just >> sporadic >> cases, but I can add an event_route to pass the buffer and some >> hints >> to configuration file in case the initial parsing fails and the >> message >> is dropped before getting to config file. >> >> It is good to know that someone use such case, introduction of >> sanity >> was clearly mentioned and so far there was no complain. >> >> Cheers, >> Daniel >> >>> On Thursday 28 October 2010, Daniel-Constantin Mierla wrote: >>>> On 10/26/10 2:22 PM, Sergey Okhapkin wrote: >>>>> setdebug() function is no longer provided by core. How to change >>>>> debug level dynamically in the script to trace execution of a few >>>>> statements only? >>>> seems was lost in transition to 3.0, I will check to see what >>>> can be >>>> done with the new architecture. >>>> >>>> Thanks, >>>> Daniel >>> _______________________________________________ >>> SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing >>> list sr-users@lists.sip-router.org >>> http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users > _______________________________________________ > SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing > list > sr-users@lists.sip-router.org > http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users _______________________________________________ SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing list sr-users@lists.sip-router.org http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users
SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing list sr-users@lists.sip-router.org http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users
The law didn't work this time, the patch works fine to me :-)
On Friday 29 October 2010, Daniel-Constantin Mierla wrote:
On 10/29/10 5:40 PM, Sergey Okhapkin wrote:
I can't test now - high volume traffic is running. But I looked at the diff, it seems to me OK, I see no reason why it could not work:-)
Because of the law, the Murphy law ;-)
Probably I will test sometime next week if you don't do it before and then backport.
Cheers, Daniel
On Friday 29 October 2010, you wrote:
On 10/29/10 4:33 PM, Daniel-Constantin Mierla wrote:
On 10/29/10 4:09 PM, Sergey Okhapkin wrote:
I just realized that 3.1 executes main routing block when broken SIP message is received! But sanity_check() function returns 0 on failed checks and script execution stops, my favorite xlog() is not executed.
OK, that can be fixed very easy. Iirc, in some cases you don't get the sip message in the config if it is really broken, but I am not sure if error_route was executed for that case anyhow.
I committed on git master, id:
39a8108c62a2beafba18417613b9660a8abef86a
Can you take the patch and test it on 3.1? If all is fine, then I can backport - even the default cfg expects non-drop behavior since it attempts to print a log message there.
Cheers, Daniel
Cheers, Daniel
Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: ERROR:<core> [parser/msg_parser.c:177]: ERROR: get_hdr_field: bad to header Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: INFO:<core> [parser/msg_parser.c:351]: ERROR: bad header field [To: Chi Nhung<sip: ] Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: ERROR:<core> [parser/msg_parser.c:177]: ERROR: get_hdr_field: bad to header Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: INFO:<core> [parser/msg_parser.c:351]: ERROR: bad header field [To: Chi Nhung<sip: ] Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: ERROR:<core> [parser/msg_parser.c:177]: ERROR: get_hdr_field: bad to header Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: INFO:<core> [parser/msg_parser.c:351]: ERROR: bad header field [To: Chi Nhung<sip: ] Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: ERROR:<core> [msg_translator.c:1946]: ERROR: build_res_buf_from_sip_req: alas, parse _headers failed Oct 29 09:48:54 west /usr/local/sbin/kamailio[12225]: WARNING: sanity [sanity.c:253]: sanity_check(): check_required_headers(): failed to s end 400 via sl reply
On Friday 29 October 2010, Daniel-Constantin Mierla wrote:
On 10/29/10 11:53 AM, Sergey Okhapkin wrote: > Daniel, I think this could be implemented without re-introduction of > error_route,
it will not be re-introduced for sure, we have now a different core framework which care reuse event_route mechanism.
In the past the error route was executed when the error was discovered, so it several places of the code, and I am not sure that the core can continue up to config execution of something is really broken.
I will have it in mind when I get some spare time.
Cheers, Daniel
> what about core parameter route_on_error=yes/no (default no)? If > route on error is enabled, then sanity module will do the trick: > > route{ > if(!sanity_check("1159")) { > xlog("L_INFO","Bad message from > $proto:$si:$sp\n$mb\n"); > break; > } > .... > > It will be up to script writer what to do in main routing block if > message parsing fails. > > On Friday 29 October 2010, Sergey Okhapkin wrote: >> It would be nice to get error_route back. The route was trivial in >> my cfg file: >> >> error_route { >> xlog("L_INFO","Bad message from $proto:$si:$sp\n$mb\n"); >> drop(); >> } >> >> On Friday 29 October 2010, Daniel-Constantin Mierla wrote: >>> On 10/28/10 1:41 PM, Sergey Okhapkin wrote: >>>> I've found one more feature missed in 3.1 - error_route is >>>> eliminated >>>> and sanity module added. But if received SIP message is >>>> malformed and >>>> can't be parsed, routing script is not executed and I have no >>>> way to >>>> log the message to kamailio log. I did it in error_route before. >>>> >>>> This logging is important to me to support customers. SIP ALG in >>>> routers often breaks Via line :-( >>> >>> Since I developed error_route, I remember that it handled just >>> sporadic >>> cases, but I can add an event_route to pass the buffer and some >>> hints >>> to configuration file in case the initial parsing fails and the >>> message >>> is dropped before getting to config file. >>> >>> It is good to know that someone use such case, introduction of >>> sanity >>> was clearly mentioned and so far there was no complain. >>> >>> Cheers, >>> Daniel >>> >>>> On Thursday 28 October 2010, Daniel-Constantin Mierla wrote: >>>>> On 10/26/10 2:22 PM, Sergey Okhapkin wrote: >>>>>> setdebug() function is no longer provided by core. How to >>>>>> change debug level dynamically in the script to trace execution >>>>>> of a few statements only? >>>>> >>>>> seems was lost in transition to 3.0, I will check to see what >>>>> can be >>>>> done with the new architecture. >>>>> >>>>> Thanks, >>>>> Daniel >>>> >>>> _______________________________________________ >>>> SIP Express Router (SER) and Kamailio (OpenSER) - sr-users >>>> mailing list sr-users@lists.sip-router.org >>>> http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users >> >> _______________________________________________ >> SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing >> list >> sr-users@lists.sip-router.org >> http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users > > _______________________________________________ > SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing > list > sr-users@lists.sip-router.org > http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users
SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing list sr-users@lists.sip-router.org http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users
-
Daniel-Constantin Mierla -
Sergey Okhapkin