I guess that I would probably implement a new ofproto provider that includes the features that you can support. You can register your new ofproto provider by inserting a call to ofproto_class_register() into ofproto_initialize(), in ofproto/ofproto.c.
On Sat, Jan 28, 2012 at 07:10:45PM +0100, Giorgio Buffa wrote: > Hello Ben, > Yes, you're definitely right! But this is not about implementing OpenFlow: > I just want to realize a management interface translation for the features > that is possible to translate (e.g. port statistics retrieval). > > Is up to the legacy network device to provide switching, routing, VLAN > tagging and so on, via hardware. Open vSwitch would just provide to me an > OpenFlow library/codebase, management framework and tools to be used by the > adaptation module. > > Kind regards, > Giorgio Buffa > > On Sat, Jan 28, 2012 at 6:49 PM, Ben Pfaff <[email protected]> wrote: > > > On Sat, Jan 28, 2012 at 06:49:36PM +0100, Giorgio Buffa wrote: > > > Two pictures with the scenario and the role of OVS in this scenario are > > > attached. Following a brief explanation. > > > > > > I would like to have a legacy network device (i.e. with no support to > > > OpenFlow) to be monitored and controlled by an OpenFlow Controller via > > the > > > OpenFlow Protocol. > > > > > > Thus, the OpenFlow Protocol is to be translated to the management > > protocol > > > exposed by the network device (e.g., SNMP) by an adaptation layer, and > > vice > > > versa. > > > > It seems unlikely to me that SNMP features exposed by any given legacy > > network device are adequate to implement OpenFlow. > > _______________________________________________ discuss mailing list [email protected] http://openvswitch.org/mailman/listinfo/discuss
