IETF
mif@jabber.ietf.org
Thursday, August 1, 2013< ^ >
danyork has set the subject to: MIF Working Group at IETF86 in Orlando
Room Configuration
Room Occupants

GMT+0
[09:44:27] maazrehan_gmail.com joins the room
[10:38:34] maazrehan_gmail.com leaves the room: Logged Out
[10:51:31] maazrehan_gmail.com joins the room
[10:55:44] maazrehan_gmail.com leaves the room: Logged Out
[10:55:58] dmitry.anipko joins the room
[10:59:36] maazrehan_gmail.com joins the room
[11:07:16] Dave Thaler joins the room
[11:07:53] <dmitry.anipko> I'm on jabber :) thanks
[11:08:04] <Dave Thaler> good morning :)
[11:09:45] <maazrehan_gmail.com> Good After noon,
[11:10:06] <dmitry.anipko> good afternoon everyone :)
[11:10:38] <maazrehan_gmail.com> I am Maaz from the mif meeting room.
[11:11:15] Ted Lemon joins the room
[11:12:30] <maazrehan_gmail.com> I hope you hear the voice.
[11:16:47] Hui Deng joins the room
[11:18:19] <dmitry.anipko> It means how some of the existing processes should be done by PVD-aware nodes
[11:18:49] <dmitry.anipko> Agreed with Dave
[11:20:55] <Dave Thaler> I said RFC 6724 separates things into sort of 2 cases
[11:22:14] <dmitry.anipko> it was Dave's comment
[11:22:19] <maazrehan_gmail.com> Yes.
[11:23:59] <dmitry.anipko> either connectbyname, or API where IP is wrapped by an object that allows to pass additional context
[11:24:04] <dmitry.anipko> with the IP
[11:24:10] <dmitry.anipko> (addition to Dave's comment)
[11:24:51] <dmitry.anipko> thanks for relaying :)
[11:25:28] <maazrehan_gmail.com> I am doing it 1st time. So I hope to improve with time.
[11:25:53] <Ted Lemon> You are doing very well.   Thank you for volunteering!
[11:26:34] <maazrehan_gmail.com> :)
[11:27:43] Victor Kuarsingh joins the room
[11:30:21] <dmitry.anipko> 2 Erik: the trust in the document is currently about trusting to definition / whether something belongs to the PVD, but not whether you trust to the results of using of each particular elements
[11:30:40] Andrew Yourtchenko joins the room
[11:35:04] <dmitry.anipko> 2 the current speaker: one of ways to establish PVD trust is by L2 information
[11:35:08] <dmitry.anipko> it is covered in the doc
[11:35:27] <dmitry.anipko> (speaker before DThaler)
[11:35:44] Dan Wing joins the room
[11:53:42] Dave Thaler leaves the room
[11:54:15] <dmitry.anipko> 2 Erik: I think it falls into discussion of design details of trust
[11:54:27] <dmitry.anipko> the arch text allows that as well as other ways
[12:00:22] <dmitry.anipko> 2 Erik: the same PVD can be advertized on train 1 and on train 2
[12:00:35] <dmitry.anipko> with prefix of train 1 on on train 1, and with prefx of train 2 on train 2
[12:03:38] <dmitry.anipko> 2 Ted: agree with scenario
[12:03:47] <dmitry.anipko> but there are different design ways to meet it
[12:03:55] <dmitry.anipko> 1 way: mutliple copies of DHCP option
[12:04:03] <dmitry.anipko> 2nd way: mlutipke copies of RA
[12:04:05] <dmitry.anipko> option
[12:04:23] <maazrehan_gmail.com> RA?
[12:04:28] <Ted Lemon> DHCP has topology int he packet, so we may elect to support two PVDs in the same DHCP packet.
[12:04:57] <dmitry.anipko> sure, that's fine too
[12:05:44] Victor Kuarsingh leaves the room
[12:07:00] <dmitry.anipko> those routers would use the same PVD ID in their explicit PVD info
[12:07:03] <dmitry.anipko> (2 Erik)
[12:07:51] <dmitry.anipko> 2 Ted: you can't re-sign, but it may be not needed, it may be possible to pass the signed record all the way down
[12:08:32] <Ted Lemon> yes, that's right.
[12:10:04] cheshire joins the room
[12:11:52] <dmitry.anipko> More typical example of app configuration in question is SIP server
[12:15:17] <dmitry.anipko> 2 Dave: the reason we're talking of PVDs is because generally merging doesn't work today
[12:15:29] <dmitry.anipko> in some specific cases merging may be possible
[12:15:43] <dmitry.anipko> but I think default safe start is that stuff is not mergable
[12:15:49] <dmitry.anipko> unless explicitly known otherwise
[12:16:18] <Ted Lemon> +1
[12:19:32] Dave Thaler joins the room
[12:24:10] <dmitry.anipko> 2 Dave: and one of the default routes points to walled garden
[12:24:27] <dmitry.anipko> another example where merge is not safe
[12:25:58] <dmitry.anipko> safe approach for routes: use them in the way such as they were different routing tables
[12:26:20] <dmitry.anipko> whether or not they are in the same hash table / list / container is just implementation specific
[12:26:50] <dmitry.anipko> preferences dont' work, walled garden may have high pref default route
[12:29:06] <dmitry.anipko> Agreed with Lorenzo, and that's what draft currently says :)
[12:29:44] <dmitry.anipko> (about ultimately being node decision, and PVDs can't directly affect each other)
[12:33:56] <maazrehan_gmail.com> .
[12:38:01] <dmitry.anipko> non-PVD apps functioning can be be improved even without merging
[12:38:05] <dmitry.anipko> (2 Dave)
[12:38:15] Dave Thaler leaves the room
[12:38:18] Victor Kuarsingh joins the room
[12:38:23] <dmitry.anipko> by OS picking one PVD
[12:38:28] <dmitry.anipko> Agree with Erik
[12:42:48] <dmitry.anipko> 2 Lorenzo: completely agree with the principles
[12:43:00] <dmitry.anipko> on implementation - starting point is implicit PVDs formed per interface
[12:43:29] <dmitry.anipko> but surely significant changed to implementations may be needed
[12:43:55] Guangqing Deng joins the room
[12:44:10] Guangqing Deng leaves the room
[12:47:54] <dmitry.anipko> should we rename to provisioning bubbles? :)
[12:49:15] <maazrehan_gmail.com> Subject changed.
[12:51:27] <Andrew Yourtchenko> pebbles :-)
[12:56:10] Dan Wing leaves the room
[12:57:17] Ted Lemon leaves the room
[12:58:54] maazrehan_gmail.com leaves the room
[12:58:58] Andrew Yourtchenko leaves the room
[12:59:35] cheshire leaves the room
[12:59:46] Victor Kuarsingh leaves the room
[13:06:54] Hui Deng leaves the room
[13:09:59] dmitry.anipko leaves the room
[13:20:16] Dan Wing joins the room
[13:34:59] Ted Lemon joins the room
[13:54:09] Dan Wing leaves the room
[13:54:15] Ted Lemon leaves the room
[13:55:09] Ted Lemon joins the room
[14:39:19] Victor Kuarsingh joins the room
[14:50:22] Victor Kuarsingh leaves the room
[15:02:05] Ted Lemon leaves the room
[15:02:09] Ted Lemon joins the room
[15:02:33] Ted Lemon leaves the room
[15:05:49] Ted Lemon joins the room
[15:10:00] Andrew Yourtchenko joins the room
[15:16:22] Ted Lemon leaves the room
[15:23:01] Andrew Yourtchenko leaves the room
[15:30:28] dmitry.anipko joins the room
[15:31:19] Ted Lemon joins the room
[15:37:07] Ted Lemon leaves the room
[15:56:13] Ted Lemon joins the room
[15:58:00] Ted Lemon leaves the room
[16:19:58] dmitry.anipko leaves the room