The OVS datapath extracts the packet headers then looks for a match in its flow table. What part of that is the "flow index"?
On Tue, Jan 06, 2015 at 02:01:51PM -0800, Ashok Chippa wrote: > Okay. How does the OVS datapath know this is a new flow? i.e. there is no > matching flow entry? Doesn't it derive some sort of an index (into the flow > table) from packet header fields? What am I missing? :) > > If this is explained in details somewhere, I would appreciate the pointer. > > Thanks, > ashok > > On Tue, Jan 6, 2015 at 1:43 PM, Ben Pfaff <[email protected]> wrote: > > > On Tue, Jan 06, 2015 at 11:01:08AM -0800, Ashok Chippa wrote: > > > Perhaps once you understand, it will make sense? :) > > > > > > Call it whatever flow_id or cookie or xyz? Let me call it "flow index" > > for > > > the sake of this discussion. The ovs datapath has some sort of a flow > > index > > > to look into the flow table when a packet arrives, right? Now, how does > > the > > > ovs datapath arrive at this flow index (upon a packet arrival)? It takes > > > certain match parameters (from the packet header and perhaps some > > > metadata) and hashes them, right? > > > > I don't think you understand how the OVS datapath works. > > _______________________________________________ discuss mailing list [email protected] http://openvswitch.org/mailman/listinfo/discuss
