> On Dec 17, 2014, at 12:09 AM, Yang Yannan <[email protected]> wrote:
>
> openflow flows support multiple tables. datapath flows comprise only a single
> table. so how openflow flows in multiple tables are transfered into datapath
> flow in a single table, especially when openflow flows contain the action
> resubmit?
The datapath is just a cache of recently seen traffic. When a packet arrives,
if it doesn't match in the cache, it is processed through the OpenFlow tables.
For example, let's say a packet matches the following OpenFlow flows:
table_id=0,in_port=1,action=resubmit(,1)
table_id=1,dl_dst=01:23:45:67:89:ab,action=3
The datpath flow can just look like the following, even though there were
multiple tables in OpenFlow:
in_port=1,dl_dst=01:23:45:67:89:ab,action=3
This is obviously a trivial example, but it illustrates the concept. Figuring
out how to generate that datapath flow can be fairly complicated, so a lot of
OVS's code is devoted to that.
--Justin
_______________________________________________
discuss mailing list
[email protected]
http://openvswitch.org/mailman/listinfo/discuss
