[ovs-dev] [PATCH v4 2/2] tunneling: Avoid recirculation on datapath by computing the recirculate actions at translate time.

[Ben Pfaff]

On Wed, Mar 08, 2017 at 08:51:38PM +0000, Chandran, Sugesh wrote:
> > From: Ben Pfaff [mailto:blp at ovn.org]
> > Sent: Tuesday, March 7, 2017 6:07 PM
> > To: Chandran, Sugesh <sugesh.chandran at intel.com>
> > Cc: dev at openvswitch.org; Gray, Mark D <mark.d.gray at intel.com>;
> > jan.scheurich at ericsson.com; jarno at ovn.org; Zoltán Balogh
> > <zoltan.balogh at ericsson.com>
> > Subject: Re: [PATCH v4 2/2] tunneling: Avoid recirculation on datapath by
> > computing the recirculate actions at translate time.
> > 
> > On Thu, Feb 09, 2017 at 03:41:54PM +0000, Sugesh Chandran wrote:
> > > Openvswitch datapath recirculates packets for tunneling, i.e.
> > > the incoming packets are encapsulated at first pass. Further actions
> > > are applied on encapsulated packets on the second pass after recirculating.
> > > The proposed patch compute and append the post tunnel actions at the
> > > time of translation itself instead of recirculating at datapath. These
> > > actions are solely depends on tunnel attributes so there is no need of
> > datapath recirculation.
> > > By avoiding the recirculation at datapath, the patch offers upto 30%
> > > performance improvement for VxLAN tunneling in our testing.
> > > The action execution logic is using the new CLONE action to define the
> > > packet cloning when the actions are combined. The lenght in the CLONE
> > > action specifies the size of nested action set.
> > >
> > > Also fixing the test suites failures that are introduced by CLONE
> > > action in tunneling except the ovn test case 2299.
> > 
> > I guess by this you mean that this test still consistently fails?
> >   2301: ovn -- 1 LR with distributed router gateway port FAILED (ovn.at:6658)
> > 
> > What's the plan for that, presumably we can't apply a patch that causes test
> > failures?
> [Sugesh] : The test case is failing due to the missing datapath rule. 
> With our limited knowledge on the OVN and OVS UT setup, Here are our findings,
> The tunnel(in the test case its geneve) create two rules in the datapath for packet forwarding.
> The proposed patch is combining them into single rule to avoid the recirculation and speed up the packet processing.
> The failing test case is trying to send packets directly to the outer bridge to match on nonexistent second rule.
> Should we comment out this output validation in the test case?
> Please let me know your inputs on this? Also looking for help from OVN experts who can suggest a fix for the same?

If I understand correctly, the now-failing check in the test case is
checking for an implementation detail that has changed.  If that's true,
then I think it would be reasonable to remove that check.