On 17 March 2016 at 08:56, Ryan Moats <[email protected]> wrote: > Guru Shetty <[email protected]> wrote on 03/17/2016 09:56:08 AM: > > > From: Guru Shetty <[email protected]> > > To: Ryan Moats/Omaha/IBM@IBMUS > > Cc: ovs dev <[email protected]> > > Date: 03/17/2016 09:56 AM > > Subject: Re: [ovs-dev] Update on OVN east-west loadbalancing. > > > > > > I gave this patch series a test drive this morning, but I'm seeing a > > failure on > > the 3HVs, 1LS, 3 lports/HV end to end OVN test case. I *think* it's > > coming from this rule: > > > > A Loadbalancer VIP is usually in a subnet different than the logical > > switch. Since I do not see a router in your test case, I can see one > > reason why Loadbalancing won't work. I would suggest starting with > > atleast one router connecting your client subnet (LS1) with the pool > > of servers in a different subnet (LS2). > > Er?!?! > > This isn't a specialized test case, this is one of the four E2E OVN > test cases that run as part of "make check" and since it doesn't have > a router, then I'd expect it should work as the LB code isn't doing > anything... >
Ah. So this is a unit test case. I will give a closer look when I re-spin the series. I had some other issues with unit tests as I was running on a specialized kernel (even without any of my patches), so I did not actually try to track down the issues. > > > Ryan (regXboi) > > > _______________________________________________ dev mailing list [email protected] http://openvswitch.org/mailman/listinfo/dev
