Hi Justin, thanks for the reply. I DO notice that OVS uses kernel route to forward the packets. But generally the host OS will only maintain host route and default route. On a server which OVS is embedded, it may be enough. But we are implementing a independent switch device with ASIC chip, which supports OVS stack and openflow, maybe we need some static LPM route to forward the encapsulated tunnel packet. Do you think so?
By the way, on the reverse direction, the de-capsulated packet will be forwarded by openflow flow entries, right? -----邮件原件----- 发件人: Justin Pettit [mailto:[email protected]] 发送时间: 2012年6月27日 10:10 收件人: Weifeng Zhang 抄送: [email protected] 主题: Re: [ovs-discuss] a GRE-Tunnel problem on the OVS and openflow enabled switch In the Linux kernel module, the packet is given to the host OS, which uses its native routing table. OVS, itself, doesn't have any knowledge of the L3 network. Presumably, if you're using OpenFlow, the system has some concept of L3 connectivity in order to connect to the controller over TCP. --Justin On Jun 26, 2012, at 10:05 PM, Weifeng Zhang wrote: > On an OVS and openflow enabled switch, GRE-tunnel is supported. Let’s assume one packet comes in and hits a flow entry, > the action is forwarded to a GRE tunnel. After GRE tunnel encapsulation, a new tunnel header is added. How should the encapsulated packet > be forwarded? By route or by manually specified tunnel outgoing port? if it’s by route, how is the route created (assume on the switch, there is no > any route protocols)? Manually create the route? > In other words, on a openflow and ovs-l2 only switch, how to forward tunnel packet? > > Weifeng Zhang, > Director of Software Engineering > > - Design Principle: Keep it Simple and Stupid > > _______________________________________________ > discuss mailing list > [email protected] > http://openvswitch.org/mailman/listinfo/discuss _______________________________________________ discuss mailing list [email protected] http://openvswitch.org/mailman/listinfo/discuss
