On Fri, Jan 22, 2010 at 11:09:52AM -0700, Brook Milligan wrote: > Manuel Bouyer writes: > > On Fri, Jan 22, 2010 at 11:26:57AM -0500, der Mouse wrote: > > that's not completely true; tap interfaces can be used without controlling > > processes. The kernel should still be able to use it as a regular > > interface for bridging purposes (it would behaves like a real > > ethernet device without the cable plugged in). > > And that is supported by the fact that it seems to work that way if > the domU interface is statically assigned. For example, ping and ssh > connections to the dom0 from the domU work fine. However, the domU > dhcp requests are not being seen by the dom0. The dom0 tcpdump > recognizes the requests on tap0 but there are no responses: > > 11:32:25.016539 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP, > Request from 00:16:3e:00:00:12 (oui Unknown), length 300 > > So, why does the bridging not work for dhcp requests when it does seem > to work for at least some other packets? I wonder if it is an issue of dhcpd not seeing incoming packets or maybe an issue of dhcpd not being able to send the answer (it uses a bpf filter for incoming packets and a raw socket for outgoing packets, if memory serves). Do you really not see anything in the logs? -- Quentin Garnier - cube%cubidou.net@localhost - cube%NetBSD.org@localhost "See the look on my face from staying too long in one place [...] every time the morning breaks I know I'm closer to falling" KT Tunstall, Saving My Face, Drastic Fantastic, 2007.
Attachment:
pgpf0K1jUqTPr.pgp
Description: PGP signature