Folks: Some of you may remember the discussion Dale and I had
on the list last week around the SIP CLF draft. I will like
to produce a summary of that discussion for those of you
who were following it.
First off, a great note of appreciation to Dale for taking his
time to provide in-depth comments. The next revision will
include much of his comments.
Next, the main area of discussion in the thread was what
constitutes a message in the context of CLF logging: is
it how rfc3261 describes it? That is, a SIP request is
one message, a response is another message, and so on?
Or does message = SIP request +
0 or more provisional responses +
1 final response
Dale was advocating uniformity in producing one line per
message (i.e., a request takes up a line, a response takes
up another line, and so on.) I believe this is a good thing.
So, on the basic issue of what constitutes a message, we are
now in agreement (thanks, RjS!), and better yet, that
agreement sticks to the rfc3261 definition of a message.
The exact nature of logging the SIP CLF elements is complex.
It seems to me that in order to reduce the complexity, we can
say that a SIP CLF is composed of:
sip-clf := request-clf / response-clf
request-clf := %d %h %u %m %r %f %t %i "%c" %x %y
response-clf := %d %x %y %s %t "%c"
Note, I have taken %s out of request-clf (i.e., we do not
put the response -- provisional or final -- in the
request-clf line. It goes in the response-clf line.)
This way, the CLF entries are common across UAS, UAC, proxies,
B2BUAs, etc. Each CLF entries for UAS and UAC would have exactly
one request-clf line (a message) and one or more
response-clf lines (messages). CLF entries for proxies and
B2BUAs can have more than one request-clf lines and more than
one response-clf lines.
I will have a new revision out sometimes next week.
Thanks,
- vijay
--
Vijay K. Gurbani, Bell Laboratories, Alcatel-Lucent
1960 Lucent Lane, Rm. 9C-533, Naperville, Illinois 60566 (USA)
Email: vkg@{alcatel-lucent.com,bell-labs.com,acm.org}
Web: http://ect.bell-labs.com/who/vkg/
_______________________________________________
Sipping mailing list https://www.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@xxxxxxxxxxxxxxx for questions on current sip
Use sip@xxxxxxxx for new developments of core SIP
