Hi Jiri, I have answered on your comments in the doc. None of the EXT-382 aspects that you debate are crucial for the L3 tunneling and NSH use cases at hand, so we shouldn't block on those.
However, I would still like to make the point that EXT-382 is designed to allow more flexibility for an OF controller to deal with encapsulation stacks. The following could be possible and make sense from an PTAP OF controller perspective to flexibly deal with stacks of VLAN tags instead of using the old "hard-wired" push/pop_vlan actions: # Ingress table add-flow table=0,in_port=1,packet_type=(0,0) actions=decap(),goto_table:10 add-flow table=0,in_port=1,packet_type=(1,0x800) actions goto_table:20 # VLAN decapsulation table add-flow table=10,packet_type=(1, 0x88A8), actions=decap(),resubmit:,10 add-flow table=10,packet_type=(1, 0x8100), actions=decap(),resubmit:,10 add-flow table=10,packet_type=(1, 0x800), actions=goto_table:20 # L3 FIB table add-flow table=20,packet_type=(1, 0x800),nw_dst=10.10.10.10 actions=dec_ttl,group:1000 add-flow table=20,packet_type=(1, 0x800),nw_dst=10.10.10.20 actions=dec_ttl,group:2000 add-flow table=20,packet_type=(1, 0x800),nw_dst=10.10.10.30 actions=dec_ttl,group:3000 # Egress groups add-group group_id=1000,type=indirect,bucket=actions=encap(0,0),set_field(dl_dst,11:22:33:44:55:66),output:2 add-group group_id=2000,type=indirect,bucket=actions=encap(1,0x8100),set_field(vlan_id,1000),group:1000 add-group group_id=3000,type=indirect,bucket=actions=encap(1,0x8100),set_field(vlan_id,1000),encap(1,0x88A8),set_field(vlan_id,333),group:1000 This is just to illustrate the point regarding encap/decap. In reality the controller would likely map the VLAN tags to metadata to match on in the FIB table for tenant VRF separation. BR, Jan > -----Original Message----- > From: Jiri Benc [mailto:jb...@redhat.com] > Sent: Thursday, 01 December, 2016 11:37 > To: Jan Scheurich <jan.scheur...@ericsson.com> > Cc: 'ben.mackcr...@corsa.com' <ben.mackcr...@corsa.com>; Zoltán Balogh > <zoltan.bal...@ericsson.com>; 'Jarno Rajahalme' > <ja...@ovn.org>; Ben Pfaff <bpf...@vmware.com>; Pravin Shelar > <pshe...@ovn.org>; Yang, Yi Y (yi.y.y...@intel.com) > <yi.y.y...@intel.com>; 'jpet...@ovn.org' <jpet...@ovn.org>; > d...@openvswitch.org > Subject: Re: Packet type-aware pipeline, L3 tunneling, EXT-382 and NSH: > Minutes call 2016-11-29 > > On Wed, 30 Nov 2016 22:52:38 +0000, Jan Scheurich wrote: > > The corresponding section in the document is already updated with the > > results of the discussion. So please have look. There's also a short > > summary of the main results at the end of the mail. > > Thanks for writing that down. I see several shortcomings in the proposal and > several things make zero sense to me as proposed > (probably > EXT-382 fault). I commented on that in the document. > > Jiri _______________________________________________ dev mailing list d...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-dev
