On Mon, Jun 14, 2021 at 3:32 AM Han Zhou <[email protected]> wrote: > > Thanks Numan and Mark. > I have only a minor comment on patch 3. I understand it may not affect > correctness for now since new datapath always triggers recompute, but I > think it's better to be explicit that the ct_zones node cannot handle it > incrementally. With that being said, for the series: > > Acked-by: Han Zhou <[email protected]>
Thanks a lot Han and Mark for the reviews. I applied this series to master with the change suggested by Han in patch 3. Thanks Numan > > On Thu, Jun 10, 2021 at 1:35 PM Mark Michelson <[email protected]> wrote: > > > Hi Numan, For patches 2-5, > > > > Acked-by: Mark Michelson <[email protected]> > > > > For patch 1, it all looks understandable to me, but I will defer to Han > > for an authoritative ACK. I'll still give my provisional: > > > > Acked-by: Mark Michelson <[email protected]> > > > > On 6/3/21 8:27 AM, [email protected] wrote: > > > From: Numan Siddique <[email protected]> > > > > > > This series splits the logical flow and physical flow processing > > > and also handles the runtime data changes for the ct zone engine data. > > > > > > New patch p2 is added in v9 to handle sbrec_chassis changes in > > > pflow_output and lflow_output engine nodes. > > > > > > The patches 2-4 added in v8 are now 3-5 in v9. > > > > > > v8 -> v9 > > > ---- > > > * Addressed review comments from Han. Removed the noop handlers > > > for some of the engine inputs. > > > * Added a new patch - p2 to handle sbrec_chassis changes in > > > pflow_output and lflow_output engine nodes. > > > > > > v7 -> v8 > > > ---- > > > * Added the ct zones I-P for datapath and runtime data changes. > > > * Removed the noop_handler for runtime data changes in pflow engine > > > node. Now there is no handler for runtime data changes for the > > > pflow engine node. > > > > > > v6 -> v7 > > > ---- > > > * Added comments on usage of noop_handler for a couple > > > of engine inputs as suggested by Han. > > > * Addressed other review comments from Han. > > > * Added the check to handle the flow changes if 'skipped_last_time' > > > is true in ofctrl_put(). > > > > > > v5 -> v6 > > > ---- > > > * Missed out checking in the uncommitted code in ofctrl.c in v4. v5 > > > fixes it. > > > * v5 accidently modified ovs submodule commit id. v6 reverts it. > > > > > > v4 -> v5 > > > ----- > > > * Addressed Han's comments. > > > > > > v3 -> v4 > > > ----- > > > * Addressed Mark G's comments. > > > * Rebased to resolve conflicts. > > > > > > v2 -> v3 > > > ----- > > > * Rebased to resolve conflicts. > > > > > > v1 -> v2 > > > ----- > > > * Rebased to resolve conflicts. > > > > > > > > > Numan Siddique (5): > > > ovn-controller: Split logical flow and physical flow processing. > > > controller: Handle sbrec_chassis changes in lflow and pflow output > > > engines. > > > ovn-controller: Handle datapath changes incrementally for ct zone I-P > > > engine node. > > > physical: Set the port binding uuid as cookie for flows where > > > relevant. > > > controller I-P: ct zone runtime data handler. > > > > > > TODO.rst | 6 + > > > controller/ofctrl.c | 99 +++-- > > > controller/ofctrl.h | 6 +- > > > controller/ovn-controller.c | 812 ++++++++++++++++++++---------------- > > > controller/physical.c | 68 ++- > > > controller/physical.h | 4 - > > > tests/ovn-performance.at | 26 ++ > > > 7 files changed, 581 insertions(+), 440 deletions(-) > > > > > > > _______________________________________________ > > dev mailing list > > [email protected] > > https://mail.openvswitch.org/mailman/listinfo/ovs-dev > > > _______________________________________________ > dev mailing list > [email protected] > https://mail.openvswitch.org/mailman/listinfo/ovs-dev > _______________________________________________ dev mailing list [email protected] https://mail.openvswitch.org/mailman/listinfo/ovs-dev
