From: Numan Siddique <num...@ovn.org> We already add below logical flows to by pass IPv6 Neighbor discovery packets from in/out ACL stage.
table=6 (ls_in_acl ), priority=65535, match=(nd), action=(next;) table=4 (ls_out_acl ), priority=65535, match=(nd), action=(next;) This patch also adds nd_rs and nd_ra to these logical flows. Without these the IPv6 Router Adv packets generated by ovn-controller are dropped if CMS has configured ACLs. Reported-by: Jakub Libosvar <jlibo...@redhat.com> Signed-off-by: Numan Siddique <num...@ovn.org> --- northd/ovn-northd.8.xml | 6 ++++++ northd/ovn-northd.c | 6 ++++-- 2 files changed, 10 insertions(+), 2 deletions(-) diff --git a/northd/ovn-northd.8.xml b/northd/ovn-northd.8.xml index 7281eeecc..a7639f33a 100644 --- a/northd/ovn-northd.8.xml +++ b/northd/ovn-northd.8.xml @@ -467,6 +467,12 @@ ACL re-allow this connection. </li> + <li> + A priority-65535 flow that allows IPv6 Neighbor solicitation, + Neighbor discover, Router solicitation and Router advertisement + packets. + </li> + <li> A priority 34000 logical flow is added for each logical switch datapath with the match <code>eth.dst = <var>E</var></code> to allow the service diff --git a/northd/ovn-northd.c b/northd/ovn-northd.c index 0fc62bf91..b8c9e9325 100644 --- a/northd/ovn-northd.c +++ b/northd/ovn-northd.c @@ -5389,8 +5389,10 @@ build_acls(struct ovn_datapath *od, struct hmap *lflows, /* Ingress and Egress ACL Table (Priority 65535). * * Not to do conntrack on ND packets. */ - ovn_lflow_add(lflows, od, S_SWITCH_IN_ACL, UINT16_MAX, "nd", "next;"); - ovn_lflow_add(lflows, od, S_SWITCH_OUT_ACL, UINT16_MAX, "nd", "next;"); + ovn_lflow_add(lflows, od, S_SWITCH_IN_ACL, UINT16_MAX, + "nd || nd_ra || nd_rs", "next;"); + ovn_lflow_add(lflows, od, S_SWITCH_OUT_ACL, UINT16_MAX, + "nd || nd_ra || nd_rs", "next;"); } /* Ingress or Egress ACL Table (Various priorities). */ -- 2.26.2 _______________________________________________ dev mailing list d...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-dev