radext - Wednesday, November 12, 2014

radext

radext@jabber.ietf.org

Wednesday, November 12, 2014< ^ >

hartmans-remote has set the subject to: IETF 90 radext meeting

Room Configuration

Room Occupants

GMT+0
[00:10:14] stefan.winter joins the room
[00:13:01] stefan.winter leaves the room
[00:27:49] stefan.winter joins the room
[00:38:17] stefan.winter leaves the room
[00:38:36] stefan.winter joins the room
[00:40:25] stefan.winter leaves the room
[00:42:23] stefan.winter joins the room
[00:55:46] stefan.winter leaves the room
[01:07:14] stefan.winter joins the room
[01:20:08] Alan DeKok joins the room
[01:21:23] Meetecho joins the room
[01:22:57] jimsch1 joins the room
[01:24:52] hartmans joins the room
[01:26:44] JoeP joins the room
[01:27:18] Parthasarathi Ravindran joins the room
[01:28:59] Alan DeKok leaves the room
[01:29:02] Parthasarathi Ravindran leaves the room
[01:29:34] Alan DeKok joins the room
[01:30:45] Alan DeKok leaves the room
[01:31:00] Alan DeKok joins the room
[01:33:39] <Alan DeKok> My $0.02 is to use IPFix for everything we can, and to define new attributes for everything else
[01:34:08] <Alan DeKok> given the interest in accounting flows, IPFix is the logical conclusion
[01:34:13] <Alan DeKok> yes
[01:35:23] <Alan DeKok> I have a draft
[01:35:51] <Alan DeKok> there's a draft to map IPFix to RADIUS TLVs
[01:36:38] <Alan DeKok> one sec
[01:36:51] <Alan DeKok> https://tools.ietf.org/html/draft-dekok-radius-ipfix-00
[01:37:10] alejandro.perez.mendez joins the room
[01:37:34] davidkazuhiro joins the room
[01:39:31] <Alan DeKok> what Sam said 100%
[01:47:42] <Alan DeKok> DNS timeouts apply to people using DNS.  Once the connection is up, you're no longer using DNS
[01:51:13] stefan.winter leaves the room
[01:56:27] stefan.winter joins the room
[01:56:34] <Alan DeKok> I have nothing to add for NAI
[02:00:46] <Alan DeKok> I don't think I have anything to add here
[02:01:09] <Alan DeKok> we'll just do the slides
[02:01:35] <hartmans> I'd rather just kill COA
[02:01:42] <Alan DeKok> sure.
[02:01:47] <Alan DeKok> people use it in production tho
[02:01:55] <Alan DeKok> Peter Deacon was VERY set on using it lots
[02:02:52] <Alan DeKok> mike?
[02:03:22] <Alan DeKok> data types should be pretty much ready to go
[02:03:44] <hartmans> Long term if it is going to stick around it would be good to address the issues noted in the document explaining why it is experimental.  That's one of the more scary this-is-half-baked statements I've ever read in an RFC
[02:04:07] <Alan DeKok> which doc is that Sam?
[02:04:12] <hartmans> COA base spec
[02:04:21] <Alan DeKok> yes, I agree
[02:14:27] <jimsch1> is identity required to be UTF-8 in RADIUS?
[02:14:45] <jimsch1> I thought part of the problem was that this was not a guarentee, just like it might not be an NAI
[02:16:10] <stefan.winter> text      Consisting only of UTF-8 encoded 10646 [7] characters.
[02:16:59] <stefan.winter> of several choices
[02:20:38] <Alan DeKok> historically people put any kind of garbage into User-Name.  Most people seem to have figured out that'
[02:20:46] <Alan DeKok> that's bad, and don't do it any more
[02:27:28] <Alan DeKok> basically if the NAI isn't UTF-8, all bets are off
[02:27:33] <Alan DeKok> and you do whatever you want to do
[02:27:39] <Alan DeKok> sure
[02:27:42] <Alan DeKok> I'm on skype
[02:27:53] <jimsch1> What about normalization?
[02:28:10] davidkazuhiro leaves the room
[02:28:30] davidkazuhiro joins the room
[02:34:49] <Alan DeKok> I support adoption of this document into RADEXT
[02:38:23] Meetecho leaves the room
[02:38:24] Meetecho joins the room
[02:41:56] Lorenzo Miniero joins the room
[02:42:08] <Alan DeKok> no one implements it
[02:42:11] <Alan DeKok> it's experimental
[02:42:20] <Alan DeKok> well.. FreeRADIUS implements it.  I don't think anyone else does
[02:42:51] Lorenzo Miniero leaves the room
[02:43:23] <Alan DeKok> TBH once we have RADIUS over TLS, we can leverage a client to server connection... why can
[02:43:32] <Alan DeKok> can't the server send CoA packets to the client?
[02:43:38] Liu Zilong joins the room
[02:43:56] <Alan DeKok> we had endless discussions about identifying a "session" in CoA.  there was no consensus
[02:45:00] jimsch1 leaves the room
[02:45:00] <Alan DeKok> thanks.
[02:45:01] JoeP leaves the room
[02:45:02] <Alan DeKok> I'll be in Dallas
[02:45:15] Alan DeKok leaves the room
[02:45:16] stefan.winter leaves the room
[02:45:20] alejandro.perez.mendez leaves the room
[02:45:45] Liu Zilong leaves the room
[02:45:55] davidkazuhiro leaves the room
[02:46:22] Meetecho leaves the room
[02:49:00] davidkazuhiro joins the room
[03:02:07] hartmans leaves the room: Disconnected: connection closed
[03:10:15] stefan.winter joins the room
[03:52:30] davidkazuhiro leaves the room
[04:03:12] stefan.winter leaves the room
[04:03:43] davidkazuhiro joins the room
[04:08:42] davidkazuhiro leaves the room
[07:11:50] stefan.winter joins the room
[07:28:45] davidkazuhiro joins the room
[07:32:31] stefan.winter leaves the room
[11:04:58] davidkazuhiro leaves the room
[15:56:54] stefan.winter joins the room
[16:54:33] stefan.winter leaves the room
[16:58:34] stefan.winter joins the room
[17:12:07] stefan.winter leaves the room
[17:35:00] stefan.winter joins the room
[17:58:11] stefan.winter leaves the room
[18:00:36] stefan.winter joins the room
[18:21:34] stefan.winter leaves the room
[18:36:52] stefan.winter joins the room
[18:47:57] stefan.winter leaves the room
[18:51:20] stefan.winter joins the room
[19:35:39] stefan.winter leaves the room
[19:48:35] davidkazuhiro joins the room
[20:25:35] hartmans joins the room
[20:25:59] hartmans leaves the room
[21:09:57] davidkazuhiro leaves the room
[21:10:14] davidkazuhiro joins the room
[21:12:59] davidkazuhiro leaves the room
[21:13:56] davidkazuhiro joins the room
[23:21:56] davidkazuhiro joins the room
[23:23:07] davidkazuhiro leaves the room
[23:23:40] davidkazuhiro leaves the room