After little more digging, I see miss upcalls and flow modify logs like below while tc flow switching to not_in_hw. I also see similar upcall while switching back to in_hw.
2025-05-21T08:14:27.189Z|19748|dpif(handler3)|DBG|system@ovs-system: miss upcall: recirc_id(0),dp_hash(0),skb_priority(0),in_port(6),skb_mark(0),ct_state(0),ct_zone(0),ct_mark(0),ct_label(0),eth(src=fa:16:7e:75:94:76,dst=00:1c:73:aa:bb:cc),eth_type(0x0800),ipv4(src=IP_A,dst=IP_B,proto=6,tos=0,ttl=64,frag=no),tcp(src=52774,dst=6835),tcp_flags(ack) tcp,vlan_tci=0x0000,dl_src=fa:16:7e:75:94:76,dl_dst=00:1c:73:aa:bb:cc,nw_src=IP_A,nw_dst=IP_B,nw_tos=0,nw_ecn=0,nw_ttl=64,nw_frag=no,tp_src=52774,tp_dst=6835,tcp_flags=ack tcp_csum:5abe 2025-05-21T08:14:27.189Z|104292|dpif(revalidator10)|DBG|system@ovs-system: put[modify] ufid:52588bba-04e5-43fa-b677-5a96dbd11f22 recirc_id(0x5),dp_hash(0/0),skb_priority(0/0),in_port(4),skb_mark(0/0),ct_state(0x22/0xf),ct_zone(0x1/0),ct_mark(0/0),ct_label(0/0),ct_tuple4(src=IP_C/ 0.0.0.0,dst=IP_D/0.0.0.0 ,proto=6/0,tp_src=58586/0,tp_dst=6821/0),eth(src=fa:16:7e:e1:50:af,dst=00:1c:73:aa:bb:cc),eth_type(0x0800),ipv4(src=IP_C/ 0.0.0.0,dst=IP_D/0.0.0.0,proto=6/0,tos=0/0,ttl=64/0,frag=no),tcp(src=58586/0,dst=6821/0),tcp_flags(0/0), actions:push_vlan(vid=111,pcp=0),1,5 On Tue, May 20, 2025 at 7:49 PM Ilya Maximets <i.maxim...@ovn.org> wrote: > On 5/20/25 4:32 PM, Serhat Rıfat Demircan wrote: > > Yes, statistics are updated. > > > > # ovs-appctl dpctl/dump-flows type=offloaded | grep dst=00:1c:73:aa:bb:cc > > > recirc_id(0x5),in_port(4),ct_state(+est-rel+rpl-inv+trk),ct_zone(0x1),ct_mark(0),eth(src=fa:16:7e:e1:50:af,dst=00:1c:73:aa:bb:cc),eth_type(0x0800),ipv4(frag=no), > packets:222, bytes:30340, used:0.170s, actions:push_vlan(vid=111,pcp=0),5 > > > recirc_id(0x5),in_port(4),ct_state(+est-rel-rpl+trk),eth(src=fa:16:7e:e1:50:af,dst=00:1c:73:aa:bb:cc),eth_type(0x0800),ipv4(frag=no), > packets:54220862, bytes:182391451911, used:0.170s, > actions:push_vlan(vid=111,pcp=0),5 > > Then one of the options would be to turn on debug logging and try > and catch the moment the flow actions change, and then correlate the > change with some other events in the log. I'd suggest to start with: > ovs-appctl vlog/set ofproto_dpif_xlate:file:dbg dpif:file:dbg > > Best regards, Ilya Maximets. > > > > > On Tue, May 20, 2025 at 3:39 PM Ilya Maximets <i.maxim...@ovn.org > <mailto:i.maxim...@ovn.org>> wrote: > > > > On 5/20/25 2:20 PM, Serhat Rıfat Demircan wrote: > > > Redirecting to br-ex happens even if there is a single hardware > offloaded port on hypervisor. Still can't find the actual reason. > > > Increasing mac-aging-time and mac-table-size did not help also. > > > > > > By the way, it is bidirectional storage(ceph) traffic. > > > > Is statistics on the datapath flows properly updated? > > i.e. if you watch 'ovs-appctl dpctl/dump-flows', do you see stats > > properly incremented when the flow is offloaded? > > > > > > > > > > > > > > On Wed, May 14, 2025 at 1:39 PM Ilya Maximets <i.maxim...@ovn.org > <mailto:i.maxim...@ovn.org> <mailto:i.maxim...@ovn.org <mailto: > i.maxim...@ovn.org>>> wrote: > > > > > > On 5/8/25 4:15 PM, Serhat Rıfat Demircan via discuss wrote: > > > > Hello Everyone, > > > > > > > > I've been testing OVS hardware offloaded ports with > Openstack for a while. > > > > I have ConnectX-5 and ConnectX-6 cards in my lab. Iperf > tests are looking > > > > promising, but when I try to use hardware offloaded ports > for krbd traffic, > > > > after some time, traffic falls back to the software path. > Another interesting > > > > thing for me is without changing anything, it can also > switch back to the > > > > hardware path. I could not find the real reason. > > > > > > > > I'm not sure if this is a configuration issue or if I'm > hitting to cards limit. > > > > If any of you guys are experienced with this all ideas are > more than welcome. > > > > > > > > Openstack version: Caracal > > > > Operation System: Ubuntu 22.04 > > > > OVS version: 3.3.0 > > > > > > > > I see tc flow change: > > > > > > > > *From Offloaded:* > > > > $ tc filter show dev ens37f0npf0vf6 ingress > > > > filter protocol ip pref 2 flower chain 0 > > > > filter protocol ip pref 2 flower chain 0 handle 0x1 > > > > dst_mac 00:1c:73:aa:bb:cc > > > > src_mac fa:16:7e:2e:8c:f0 > > > > eth_type ipv4 > > > > ip_flags nofrag > > > > skip_sw > > > > in_hw in_hw_count 1 > > > > action order 1: vlan push id 111 protocol 802.1Q priority 0 > pipe > > > > index 4 ref 1 bind 1 > > > > no_percpu > > > > used_hw_stats delayed > > > > > > > > action order 2: mirred (Egress Redirect to device openstack) > stolen > > > > index 4 ref 1 bind 1 > > > > cookie 605b82e5b94ef3fa63c3e78349851a27 > > > > no_percpu > > > > used_hw_stats delayed > > > > > > > > *To software path:* > > > > $ tc filter show dev ens37f0npf0vf6 ingress > > > > filter protocol ip pref 2 flower chain 0 > > > > filter protocol ip pref 2 flower chain 0 handle 0x1 > > > > dst_mac 00:1c:73:aa:bb:cc > > > > src_mac fa:16:7e:2e:8c:f0 > > > > eth_type ipv4 > > > > ip_flags nofrag > > > > not_in_hw > > > > action order 1: vlan push id 111 protocol 802.1Q priority 0 > pipe > > > > index 4 ref 1 bind 1 > > > > no_percpu > > > > > > > > action order 2: skbedit ptype host pipe > > > > index 11 ref 1 bind 1 > > > > > > > > action order 3: mirred (Ingress Mirror to device br-ex) pipe > > > > index 13 ref 1 bind 1 > > > > cookie 5037716e444e83b81c4ef9844df2377d > > > > no_percpu > > > > > > > > action order 4: mirred (Egress Redirect to device openstack) > stolen > > > > index 15 ref 1 bind 1 > > > > cookie 5037716e444e83b81c4ef9844df2377d > > > > no_percpu > > > > > > Here you can see that traffic is getting broadcasted on all > ports in the > > > br-ex bridge, and since some of these ports are virtual, this > operation > > > cannot be offloaded to hardware. > > > > > > I would guess, after some time the FDB entry expires and OVS > no longer > > > knows on which port the traffic supposed to be sent out. Is > that a > > > unidirection traffic? i.e. are you also sending traffic in > the opposite > > > direction? If not, are you sending some periodic ARP requests > for the > > > bridge to learn where the destination is? > > > > > > If the traffic is bidirectional, then it might be a NIC > driver/firmware > > > issue not providing updated stats. > > > > > > Best regards, Ilya Maximets. > > > > > > >
_______________________________________________ discuss mailing list disc...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-discuss
