During my attempts to get the latest bridge networking model working on my system, I often invoked ''network-bridge stop'' with total failure. Attached is an updated version that works. -Andrew signed-off-by: Andrew Theurer <habanero@us.ibm.com> _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
On 14 Sep 2005, at 16:51, Andrew Theurer wrote:> During my attempts to get the latest bridge networking model working > on my > system, I often invoked ''network-bridge stop'' with total failure. > Attached > is an updated version that works.Just to clarify: presumably the problem with the current script, that your patch fixes, is that it tests for and tears down veth0, when in fact veth0 has been renamed to eth0, and eth0 to peth0? :-) -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Keir Fraser wrote:> > On 14 Sep 2005, at 16:51, Andrew Theurer wrote: > >> During my attempts to get the latest bridge networking model working >> on my >> system, I often invoked ''network-bridge stop'' with total failure. >> Attached >> is an updated version that works. > > > Just to clarify: presumably the problem with the current script, that > your patch fixes, is that it tests for and tears down veth0, when in > fact veth0 has been renamed to eth0, and eth0 to peth0? :-)Yes, the script in current xen-unstable, when running "stop", tests for veth0, which does not exist because it has been renamed to eth0. It now tests for peth0, since that is the only real indication that veth0 was used in the first place. There are other small fixes like removal or hard coded "peth0", removal of unconditional ''brctl delif xen-br0 eth0'', and reversal of ethernet naming. -Andrew _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
> Yes, the script in current xen-unstable, when running "stop", > tests for veth0, which does not exist because it has been > renamed to eth0. It now tests for peth0, since that is the > only real indication that veth0 was > used in the first place. There are other small fixes like > removal or > hard coded "peth0", removal of unconditional ''brctl delif > xen-br0 eth0'', and reversal of ethernet naming.Thanks for tidying that up -- its been on my todo list. Putting the networking back as it was is quite a dangerous game, so its not something we want to invoke automatically. BTW: it might be worth experimenting with inserting ''ip route cache flush'' runes in a couple of the places in the script to avoid the stall in ssh connections when invoking network-bridge start. Thanks, Ian _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel