mboned@conference.ietf.jabber.com - 2003/03/17
[12:16] %% joelja has arrived.
[12:20] %% joelja has left.
[12:58] %% jtk has arrived.
[13:40] %% jtk has left.
[14:24] %% mjh has arrived.
[14:25] %% Bill Fenner has arrived.
[14:28] <Bill Fenner> I am in OSPF with my AD hat on, what's going on in mboned?
[14:30] %% mrose has arrived.
[14:35] %% gbourdon has arrived.
[14:42] %% alex has arrived.
[14:50] %% mattb has arrived.
[14:52] %% jtk has arrived.
[14:52] <jtk> i assume I missed the DoS attacks portion?
[14:56] %% Isomer has arrived.
[14:59] %% louie has arrived.
[15:00] %% stephennadas has arrived.
[15:02] %% carlalf has arrived.
[15:02] <Bill Fenner> my high level summary of the DoS attack portion: PIM-SM and MSDP are the problem; SSM is a potential solution, since MSDP is hard to filter.
[15:02] <Bill Fenner> Dino is proposing more types of filtering in MSDP, and asking if SSM requires IGMPv3 or if there's a way to get around the host<>router protocol issue
[15:03] <Bill Fenner> but the discussion is going way too quickly for a blow by blow
[15:03] <jtk> no mention of IGMP pull style attacks? not that is the major concern, but... a concern
[15:04] <Bill Fenner> Bill Nickless agrees that there's an IGMPv3 deployment problem, especially e.g. snooping switches
[15:05] <Bill Fenner> no mention of IGMP pulling; I think the general feeling on that is that an end system can also ask a streaming server to send it data, so it's not a different attack than is possible with unicast
[15:06] <jtk> except that igmp pulls cause additional middle-box state, unicast doesn't
[15:06] <Bill Fenner> indeed
[15:06] <Bill Fenner> Slammer/Sapphire worm discussion now
[15:08] %% mrose has left.
[15:09] %% stephennadas has left.
[15:10] <Isomer> Anything interesting come up in that section?
[15:13] %% mrose has arrived.
[15:15] <jtk> would having router vendors not forward igmp past an interface by default be acceptable to this group? That (and other IP protocols) are often used in remote DoS attacks) and I don't seem to recall a need to actually have IGMP routed
[15:17] <Bill Fenner> worm summary: the worm was not targeted at multicast and almost took multicast down; something that is targeted at multicast can easily take it out
[15:18] <Bill Fenner> and the summary is basically just like the previous one, that ASM is hard to protect against DOS attacks
[15:18] <Isomer> Cool.
[15:26] %% JavierA has arrived.
[15:27] %% JavierA has left.
[15:30] <Isomer> what's the current point of discussion?
[15:38] <Bill Fenner> MCOP, but I've been out of the room
[15:39] <Isomer> hmm, fair enough
[15:44] <mjh> Next up: Ipv4 to IPv6 gateway
[15:47] <Isomer> for multicast?
[15:50] <mjh> yes
[15:51] %% mdf has arrived.
[15:53] <mdf> is stig presenting now? (on his gateway)
[15:53] %% gbourdon has left.
[15:57] <mdf> was anyone in the network conf bof this morning?
[15:58] <mrose> yes.
[15:58] <mdf> what was the upshot? forming a wg?
[15:58] <mrose> my guess is "yes", but randy asked for a draft charter to look at, and they didn't have one ready for him...
[16:00] <mdf> any feedback on the gateway draft? (in mboned)
[16:05] <mdf> I guess not
[16:06] <mjh> yes, but too rapid to summarise
[16:10] %% carlalf has left.
[16:11] <mdf> ok - thanks
[16:18] %% mrose has left.
[16:18] %% alex has left.
[16:20] %% louie has left.
[16:36] %% mdf has left.
[16:37] %% Bill Fenner has left.
[16:38] %% mjh has left.
[16:47] %% jtk has left.
[17:02] %% Bill Fenner has arrived.
[17:07] %% Bill Fenner has left.
[17:16] %% Isomer has left.
[17:52] %% venaas has arrived.
[17:53] %% venaas has left.
[23:20] %% mattb has left.