Yes, you probably need to create a new PR for that, against git master. For 5.8
branch it will be cherry-picked.
--
Reply to this email directly or view it on GitHub:
https://github.com/kamailio/kamailio/pull/3699#issuecomment-1988117447
You are receiving this because you are subscribed to this
Hi, @miconda !
About the docs changes above, do i need to make a new PR, this one is closed ?
--
Reply to this email directly or view it on GitHub:
https://github.com/kamailio/kamailio/pull/3699#issuecomment-1988089752
You are receiving this because you are subscribed to this thread.
Message
Thansk for merging @miconda !
I wanted to offer a version of the documentation for flag 16 to make it more
clear:
If bit five is set, then try to find the most closest target from all
dispatcher targets with the mandatory ip and combination of local socket,
protocol, port.
The weighted search
Merged #3699 into master.
--
Reply to this email directly or view it on GitHub:
https://github.com/kamailio/kamailio/pull/3699#event-11902601725
You are receiving this because you are subscribed to this thread.
Message ID: ___
Kamailio (SER) -
I am merging it, then rename the option for FULLADDRSOCK trying to reflect
better that socket alone has more priority than port+protocol.
--
Reply to this email directly or view it on GitHub:
https://github.com/kamailio/kamailio/pull/3699#issuecomment-1961172542
You are receiving this because
Hi !
Yes, i need to refine the docs, will try to do at this weekends.
--
Reply to this email directly or view it on GitHub:
https://github.com/kamailio/kamailio/pull/3699#issuecomment-1957261250
You are receiving this because you are subscribed to this thread.
Message ID:
@Den4t @miconda Any more questions or things that needs to be discussed?
Otherwise its probably a good idea to merge it in the next days before the
window for 5.8.0 closes completely.
--
Reply to this email directly or view it on GitHub:
> Your last comment does not clarify if a destination record matching (ip,
> socket) has to be selected over one that matches (ip, port, proto). Can you
> present the order of selection based on matching the values for ip, port,
> proto or socket?
The strictness are defined by the flags
Your last comment does not clarify if a destination record matching (ip,
socket) has to be selected over one that matches (ip, port, proto). Can you
present the order of selection based on matching the values for ip, port, proto
or socket?
--
Reply to this email directly or view it on GitHub:
Hi!
Sorry, I was sick last week...
> because of the flag for socket matching, record matching (ip, socket) is
> selected. Is this the desired behavior.
Yes, this is the desired behavior, because we can only have a target with an ip
and socket like this (it seems to be a valid configuration):
Somehow I got it that the destination that has the most matches is selected. As
we have ip. port, protocol and socket, I thought that if a record matches three
attributes (ip, port, protocol) is selected over another record that matches
two attributes (ip, socket). But as I could understand
> I meant for the case/flag DS_MATCH_FULLADDRSOCK: a destination record that
> matches **only** the socket is selected against one that matches
> address+port+proto.
Using DS_MATCH_TRY_FULL ADDR SOCK, at least an ip must also be matched, and in
this mode we do not interrupt the serach cycle,
I meant for the case/flag DS_MATCH_FULLADDRSOCK: a destination record that
matches **only** the socket is selected against one that matches
address+port+proto.
--
Reply to this email directly or view it on GitHub:
https://github.com/kamailio/kamailio/pull/3699#issuecomment-1888669573
You are
> If I understand the code correctly, a destination record that matches only
> the socket is selected against one that matches address/port/proto? Is it
> like that and if yes, is it the expected result?
Yes, it is a separate flag - DS_MATCH_SOCKET (8), it work like
ds_is_from_list(-1,0), but
If I understand the code correctly, a destination record that matches only the
socket is selected against one that matches address/port/proto? Is it like that
and if yes, is it the expected result?
--
Reply to this email directly or view it on GitHub:
I have changed the flag name and added docs to xml.
--
Reply to this email directly or view it on GitHub:
https://github.com/kamailio/kamailio/pull/3699#issuecomment-1884558285
You are receiving this because you are subscribed to this thread.
Message ID:
@Den4t pushed 1 commit.
53ae14294f021ae615c0fae6ec9b48758d9a051e dispatcher: added two new flags to
mode parameter of ds_is_from_list function for more strictly matching
--
View it on GitHub:
> I've not done any testing myself, but by my understanding the
> DS_MATCH_STRICTEST flag is basically a flag to indicate it should not return
> on the first match but scan through all. Then sort and return the one that is
> matching one most criterias?
Right, but not exactly, there is no
> Maybe DS_MATCH_FULLADDRSOCK.
OK, i will change the flag name, need some time to think about adequate name,
writing docs hope can help with this.
--
Reply to this email directly or view it on GitHub:
https://github.com/kamailio/kamailio/pull/3699#issuecomment-1879191053
You are receiving
I've not done any testing myself, but by my understanding the
DS_MATCH_STRICTEST flag is basically a flag to indicate it should not return on
the first match but scan through all. Then sort and return the one that is
matching one most criterias?
--
Reply to this email directly or view it on
Thanks for the PR!
I haven't done a full review, but first, you have to add documentation to the
xml fine in the doc/ subfolder. It helps to understand how is supposed to be
used and see if follows the code.
Also, the DS_MATCH_STRICTEST should be changed to try to reflect better what
kind of
21 matches
Mail list logo
