On Mon, Jul 12, 2010 at 02:57:51PM +0200, Manuel Franceschini wrote: > On Mon, Jul 12, 2010 at 1:57 PM, Iustin Pop <[email protected]> wrote: > > On Fri, Jul 09, 2010 at 04:58:50PM +0200, Manuel Franceschini wrote: > >> This patch series basically adds a new parameter 'family' to the > >> constructors > >> of daemon.AsyncUDPSocket and confd.client.ConfdUDPClient. This enables the > >> users of these two classes to support IPv6. > >> > >> In ganeti-confd.ConfdAsyncUDPClient a method to check the address families > >> of > >> all peers is added. > >> > >> Furthermore it adds unittests for the added functionality. > >> --- > >> I left the explicit for loop in > >> confd.client.ConfdClient._SetPeersAddressFamily() as my timeit's show > >> that it is slightly faster than the set variant and gives access to the > >> peer that cause a potential exception. > >> > >> For now it adds a TODO in ganeti-confd about a future address family > >> check. > >> > > LGTM, with one comment: > > > >> @@ -331,7 +332,12 @@ class AsyncUDPSocket(GanetiBaseAsyncoreDispatcher): > >> constants.MAX_UDP_DATA_SIZE) > >> if recv_result is not None: > >> payload, address = recv_result > >> - ip, port = address > >> + if self._family == socket.AF_INET6: > >> + # we ignore 'flow info' and 'scope id' as we don't need it > > > > s/it/them/ > > Done. > > > And the 10.0.0.x addresses should really be changed to the test address > > range (not the private one). > > I'll take the LGTM for this one and submit a patch with that change.
SGTM, thanks. iustin
