Please don't drop the mailing list. It's hard for me to guess whether this is an issue with ovs-vswitchd or something wrapped around it. To tell the difference, I would start by doing a lower-level start. One way to do that is to run "ps ax|grep ovs-vswitchd" to see the ovs-vswitchd command line, then kill ovs, then restart it by running the full command line you saw before. Another way, that will work if you see --monitor in the OVS command line, is to just kill ovs-vswitchd with a SEGV signal, e.g. "kill -SEGV `cat /var/run/openvswitch/ovs-vswitchd.pid`".
If you restart OVS one of those ways, do you still see the issue? On Thu, Jan 28, 2016 at 04:51:45PM -0600, EDGAR CANTU wrote: > > I'm using "systemctl restart openvswitch" > This is on Centos 7.1 > > > > > From: Ben Pfaff <b...@ovn.org> > To: EDGAR CANTU/Austin/IBM@IBMUS > Cc: b...@openvswitch.org > Date: 01/28/2016 04:22 PM > Subject: Re: [ovs-discuss] restarting openvswitch leaves ovs bridge down > and with no IP > > > > On Wed, Jan 27, 2016 at 04:04:49PM -0600, EDGAR CANTU wrote: > > The openvswitch version on master seems to be wiping out the bridge's IP > > when openvswitch gets restarted and it is only after restarting network > > services that the IP comes back. > > I'm using commit 6872bf1eb7e46f9c202d512aa5ca6d99fe35d23e > > I've also verified that I don' see this issue with versions 2.3.1 and > 2.4.0 > > What are you doing to restart Open vSwitch? > > > _______________________________________________ discuss mailing list discuss@openvswitch.org http://openvswitch.org/mailman/listinfo/discuss
