> Hi Lorenzo,
Hi Mark,
thx for the review.
>
> The code looks fine to me, but I'm a bit confused by the new test.
>
> My understanding of the new feature is that each ECMP nexthop has an ID
> associated with it. This ID gets placed in the ct.label. If the ECMP route
> is removed, then we can find the associated ID, and remove the conntrack
> entry that has this ID in its label.
correct.
>
> In the test, the setup initially seems good. Alice is set up as a server,
> and Bob is set up as a client behind two routers. Bob sends traffic via R2
> and we can see the conntrack entry that ensures traffic from Alice will get
> routed to R2 instead of R3. So far so good.
>
> After this point, I'm a bit confused. I expected something like:
> * Remove the ECMP route from R1 to R2 (or alter its nexthop address)
I remove all the static routes in R1 at the end of the IPv{4,6} test and then
I check we do not have any related entry in the ct table.
$ovn-nbctl --wait=hv lr-route-del R1
AT_CHECK([ovs-appctl dpctl/dump-conntrack | FORMAT_CT(172.16.0.1)])
Am I missing something?
I change the R2_ext mac address during the test, maybe we can get rid of the ct
flushing there, but I guess it is just a leftover of the previous test.
> * Check conntrack to ensure the existing entry is removed.
> * If possible, send new traffic from Bob to Alice and ensure that a new
> conntrack entry is created to show the updated ECMP symmetric reply.
We do it or am I missing something?
Regards,
Lorenzo
>
> Instead, the ECMP routes are never changed. Other configuration changes are
> made, and conntrack is flushed between each stage of the test. Am I missing
> something here?
>
> On 3/12/24 11:59, Lorenzo Bianconi wrote:
> > Introduce ecmp_nexthop_monitor in ovn-controller in order to track and
> > flush ecmp-symmetric reply ct entires when requested by the CMS (e.g
> > removing the related static routes).
> >
> > Signed-off-by: Lorenzo Bianconi <lorenzo.bianc...@redhat.com>
> > ---
> > controller/ofctrl.c | 101 ++++++++++++++
> > controller/ofctrl.h | 2 +
> > controller/ovn-controller.c | 2 +
> > tests/system-ovn-kmod.at | 266 ++++++++++++++++++++++++++++++++++++
> > tests/system-ovn.at | 4 +
> > 5 files changed, 375 insertions(+)
> >
> > diff --git a/controller/ofctrl.c b/controller/ofctrl.c
> > index f14cd79a8..a10b0f567 100644
> > --- a/controller/ofctrl.c
> > +++ b/controller/ofctrl.c
> > @@ -388,9 +388,24 @@ struct meter_band_entry {
> > static struct shash meter_bands;
> > +static struct hmap ecmp_nexthop_map;
> > +struct ecmp_nexthop_entry {
> > + struct hmap_node node;
> > + bool erase;
> > +
> > + char *nexthop;
> > + int id;
> > +};
> > +
> > static void ofctrl_meter_bands_destroy(void);
> > static void ofctrl_meter_bands_clear(void);
> > +static void ecmp_nexthop_monitor_destroy(void);
> > +static void ecmp_nexthop_monitor_run(
> > + const struct sbrec_ecmp_nexthop_table *enh_table,
> > + struct ovs_list *msgs);
> > +
> > +
> > /* MFF_* field ID for our Geneve option. In S_TLV_TABLE_MOD_SENT, this is
> > * the option we requested (we don't know whether we obtained it yet). In
> > * S_CLEAR_FLOWS or S_UPDATE_FLOWS, this is really the option we have. */
> > @@ -429,6 +444,7 @@ ofctrl_init(struct ovn_extend_table *group_table,
> > groups = group_table;
> > meters = meter_table;
> > shash_init(&meter_bands);
> > + hmap_init(&ecmp_nexthop_map);
> > }
> >
> > /* S_NEW, for a new connection.
> > @@ -883,6 +899,7 @@ ofctrl_destroy(void)
> > expr_symtab_destroy(&symtab);
> > shash_destroy(&symtab);
> > ofctrl_meter_bands_destroy();
> > + ecmp_nexthop_monitor_destroy();
> > }
> > uint64_t
> > @@ -2306,6 +2323,87 @@ add_meter(struct ovn_extend_table_info *m_desired,
> > ofctrl_meter_bands_alloc(sb_meter, m_desired, msgs);
> > }
> > +static void
> > +ecmp_nexthop_monitor_free_entry(struct ecmp_nexthop_entry *e,
> > + struct ovs_list *msgs)
> > +{
> > + if (msgs) {
> > + ovs_u128 mask = {
> > + /* ct_labels.label BITS[96-127] */
> > + .u64.hi = 0xffffffff00000000,
> > + };
> > + uint64_t id = e->id;
> > + ovs_u128 nexthop = {
> > + .u64.hi = id << 32,
> > + };
> > + struct ofp_ct_match match = {
> > + .labels = nexthop,
> > + .labels_mask = mask,
> > + };
> > + struct ofpbuf *msg = ofp_ct_match_encode(&match, NULL,
> > +
> > rconn_get_version(swconn));
> > + ovs_list_push_back(msgs, &msg->list_node);
> > + }
> > + free(e->nexthop);
> > + free(e);
> > +}
> > +
> > +static void
> > +ecmp_nexthop_monitor_destroy(void)
> > +{
> > + struct ecmp_nexthop_entry *e;
> > + HMAP_FOR_EACH_POP (e, node, &ecmp_nexthop_map) {
> > + ecmp_nexthop_monitor_free_entry(e, NULL);
> > + }
> > + hmap_destroy(&ecmp_nexthop_map);
> > +}
> > +
> > +static struct ecmp_nexthop_entry *
> > +ecmp_nexthop_monitor_lookup(char *nexthop)
> > +{
> > + uint32_t hash = hash_string(nexthop, 0);
> > + struct ecmp_nexthop_entry *e;
> > +
> > + HMAP_FOR_EACH_WITH_HASH (e, node, hash, &ecmp_nexthop_map) {
> > + if (!strcmp(e->nexthop, nexthop)) {
> > + return e;
> > + }
> > + }
> > + return NULL;
> > +}
> > +
> > +static void
> > +ecmp_nexthop_monitor_run(const struct sbrec_ecmp_nexthop_table *enh_table,
> > + struct ovs_list *msgs)
> > +{
> > + struct ecmp_nexthop_entry *e;
> > + HMAP_FOR_EACH (e, node, &ecmp_nexthop_map) {
> > + e->erase = true;
> > + }
> > +
> > + const struct sbrec_ecmp_nexthop *sbrec_ecmp_nexthop;
> > + SBREC_ECMP_NEXTHOP_TABLE_FOR_EACH (sbrec_ecmp_nexthop, enh_table) {
> > + e = ecmp_nexthop_monitor_lookup(sbrec_ecmp_nexthop->nexthop);
> > + if (!e) {
> > + e = xzalloc(sizeof *e);
> > + e->nexthop = xstrdup(sbrec_ecmp_nexthop->nexthop);
> > + e->id = sbrec_ecmp_nexthop->id;
> > + uint32_t hash = hash_string(e->nexthop, 0);
> > + hmap_insert(&ecmp_nexthop_map, &e->node, hash);
> > + } else {
> > + e->erase = false;
> > + }
> > + }
> > +
> > + HMAP_FOR_EACH_SAFE (e, node, &ecmp_nexthop_map) {
> > + if (e->erase) {
> > + hmap_remove(&ecmp_nexthop_map, &e->node);
> > + ecmp_nexthop_monitor_free_entry(e, msgs);
> > + }
> > + }
> > +
> > +}
> > +
> > static void
> > installed_flow_add(struct ovn_flow *d,
> > struct ofputil_bundle_ctrl_msg *bc,
> > @@ -2664,6 +2762,7 @@ ofctrl_put(struct ovn_desired_flow_table *lflow_table,
> > struct shash *pending_ct_zones,
> > struct hmap *pending_lb_tuples,
> > struct ovsdb_idl_index *sbrec_meter_by_name,
> > + const struct sbrec_ecmp_nexthop_table *enh_table,
> > uint64_t req_cfg,
> > bool lflows_changed,
> > bool pflows_changed)
> > @@ -2704,6 +2803,8 @@ ofctrl_put(struct ovn_desired_flow_table *lflow_table,
> > /* OpenFlow messages to send to the switch to bring it up-to-date. */
> > struct ovs_list msgs = OVS_LIST_INITIALIZER(&msgs);
> > + ecmp_nexthop_monitor_run(enh_table, &msgs);
> > +
> > /* Iterate through ct zones that need to be flushed. */
> > struct shash_node *iter;
> > SHASH_FOR_EACH(iter, pending_ct_zones) {
> > diff --git a/controller/ofctrl.h b/controller/ofctrl.h
> > index 502c73da6..e08b354f4 100644
> > --- a/controller/ofctrl.h
> > +++ b/controller/ofctrl.h
> > @@ -31,6 +31,7 @@ struct ofpbuf;
> > struct ovsrec_bridge;
> > struct ovsrec_open_vswitch_table;
> > struct sbrec_meter_table;
> > +struct sbrec_ecmp_nexthop_table;
> > struct shash;
> > struct ovn_desired_flow_table {
> > @@ -59,6 +60,7 @@ void ofctrl_put(struct ovn_desired_flow_table
> > *lflow_table,
> > struct shash *pending_ct_zones,
> > struct hmap *pending_lb_tuples,
> > struct ovsdb_idl_index *sbrec_meter_by_name,
> > + const struct sbrec_ecmp_nexthop_table *enh_table,
> > uint64_t nb_cfg,
> > bool lflow_changed,
> > bool pflow_changed);
> > diff --git a/controller/ovn-controller.c b/controller/ovn-controller.c
> > index 1c9960c70..28cac2683 100644
> > --- a/controller/ovn-controller.c
> > +++ b/controller/ovn-controller.c
> > @@ -5945,6 +5945,8 @@ main(int argc, char *argv[])
> > &ct_zones_data->pending,
> > &lb_data->removed_tuples,
> > sbrec_meter_by_name,
> > + sbrec_ecmp_nexthop_table_get(
> > + ovnsb_idl_loop.idl),
> > ofctrl_seqno_get_req_cfg(),
> > engine_node_changed(&en_lflow_output),
> > engine_node_changed(&en_pflow_output));
> > diff --git a/tests/system-ovn-kmod.at b/tests/system-ovn-kmod.at
> > index 14fe4ecec..88c196874 100644
> > --- a/tests/system-ovn-kmod.at
> > +++ b/tests/system-ovn-kmod.at
> > @@ -1054,3 +1054,269 @@ OVS_TRAFFIC_VSWITCHD_STOP(["
> > "])
> > AT_CLEANUP
> > ])
> > +
> > +OVN_FOR_EACH_NORTHD([
> > +AT_SETUP([ECMP symmetric reply - kmod])
> > +AT_KEYWORDS([ecmp])
> > +
> > +CHECK_CONNTRACK()
> > +ovn_start
> > +
> > +OVS_TRAFFIC_VSWITCHD_START()
> > +ADD_BR([br-int])
> > +
> > +# Set external-ids in br-int needed for ovn-controller
> > +ovs-vsctl \
> > + -- set Open_vSwitch . external-ids:system-id=hv1 \
> > + -- set Open_vSwitch .
> > external-ids:ovn-remote=unix:$ovs_base/ovn-sb/ovn-sb.sock \
> > + -- set Open_vSwitch . external-ids:ovn-encap-type=geneve \
> > + -- set Open_vSwitch . external-ids:ovn-encap-ip=169.0.0.1 \
> > + -- set bridge br-int fail-mode=secure
> > other-config:disable-in-band=true
> > +
> > +# Start ovn-controller
> > +start_daemon ovn-controller
> > +
> > +# Logical network:
> > +# Alice is connected to gateway router R1. R1 is connected to two
> > "external"
> > +# routers, R2 and R3 via an "ext" switch.
> > +# Bob is connected to both R2 and R3. R1 contains two ECMP routes, one
> > through R2
> > +# and one through R3, to Bob.
> > +#
> > +# alice -- R1 -- ext ---- R2
> > +# | \
> > +# | bob
> > +# | /
> > +# + ----- R3
> > +#
> > +# For this test, Bob sends request traffic through R2 to Alice. We want to
> > ensure that
> > +# all response traffic from Alice is routed through R2 as well.
> > +
> > +ovn-nbctl create Logical_Router name=R1 options:chassis=hv1
> > +ovn-nbctl create Logical_Router name=R2
> > +ovn-nbctl create Logical_Router name=R3
> > +
> > +ovn-nbctl ls-add alice
> > +ovn-nbctl ls-add bob
> > +ovn-nbctl ls-add ext
> > +
> > +# connect alice to R1
> > +ovn-nbctl lrp-add R1 alice 00:00:01:01:02:03 10.0.0.1/24 fd01::1/64
> > +ovn-nbctl lsp-add alice rp-alice -- set Logical_Switch_Port rp-alice \
> > + type=router options:router-port=alice addresses='"00:00:01:01:02:03"'
> > +
> > +# connect bob to R2
> > +ovn-nbctl lrp-add R2 R2_bob 00:00:02:01:02:03 172.16.0.2/16 fd07::2/64
> > +ovn-nbctl lsp-add bob rp2-bob -- set Logical_Switch_Port rp2-bob \
> > + type=router options:router-port=R2_bob addresses='"00:00:02:01:02:03"'
> > +
> > +# connect bob to R3
> > +ovn-nbctl lrp-add R3 R3_bob 00:00:02:01:02:04 172.16.0.3/16 fd07::3/64
> > +ovn-nbctl lsp-add bob rp3-bob -- set Logical_Switch_Port rp3-bob \
> > + type=router options:router-port=R3_bob addresses='"00:00:02:01:02:04"'
> > +
> > +# Connect R1 to ext
> > +ovn-nbctl lrp-add R1 R1_ext 00:00:04:01:02:03 20.0.0.1/24 fd02::1/64
> > +ovn-nbctl lsp-add ext r1-ext -- set Logical_Switch_Port r1-ext \
> > + type=router options:router-port=R1_ext addresses='"00:00:04:01:02:03"'
> > +
> > +# Connect R2 to ext
> > +ovn-nbctl lrp-add R2 R2_ext 00:00:04:01:02:04 20.0.0.2/24 fd02::2/64
> > +ovn-nbctl lsp-add ext r2-ext -- set Logical_Switch_Port r2-ext \
> > + type=router options:router-port=R2_ext addresses='"00:00:04:01:02:04"'
> > +
> > +# Connect R3 to ext
> > +ovn-nbctl lrp-add R3 R3_ext 00:00:04:01:02:05 20.0.0.3/24 fd02::3/64
> > +ovn-nbctl lsp-add ext r3-ext -- set Logical_Switch_Port r3-ext \
> > + type=router options:router-port=R3_ext addresses='"00:00:04:01:02:05"'
> > +
> > +# Install ECMP routes for alice.
> > +ovn-nbctl --ecmp-symmetric-reply --policy="src-ip" lr-route-add R1
> > 10.0.0.0/24 20.0.0.2
> > +ovn-nbctl --ecmp-symmetric-reply --policy="src-ip" lr-route-add R1
> > 10.0.0.0/24 20.0.0.3
> > +
> > +# Static Routes
> > +ovn-nbctl lr-route-add R2 10.0.0.0/24 20.0.0.1
> > +ovn-nbctl lr-route-add R3 10.0.0.0/24 20.0.0.1
> > +
> > +# Logical port 'alice1' in switch 'alice'.
> > +ADD_NAMESPACES(alice1)
> > +# Only send 1 router solicitation as any additional ones can cause datapath
> > +# flows to get evicted, causing unexpected failures below.
> > +NS_CHECK_EXEC([alice1], [sysctl -w
> > net.ipv6.conf.default.router_solicitations=1], [0], [dnl
> > +net.ipv6.conf.default.router_solicitations = 1
> > +])
> > +ADD_VETH(alice1, alice1, br-int, "10.0.0.2/24", "f0:00:00:01:02:04", \
> > + "10.0.0.1")
> > +NS_CHECK_EXEC([alice1], [ip -6 addr add fd01::2/64 dev alice1 nodad])
> > +NS_CHECK_EXEC([alice1], [ip -6 route add default via fd01::1])
> > +NS_CHECK_EXEC([alice1], [ip -6 neigh add fd01::1 lladdr 00:00:01:01:02:03
> > dev alice1], [0])
> > +ovn-nbctl lsp-add alice alice1 \
> > +-- lsp-set-addresses alice1 "f0:00:00:01:02:04 10.0.0.2 fd01::2"
> > +
> > +# Logical port 'bob1' in switch 'bob'.
> > +ADD_NAMESPACES(bob1)
> > +# Only send 1 router solicitation as any additional ones can cause datapath
> > +# flows to get evicted, causing unexpected failures below.
> > +NS_CHECK_EXEC([bob1], [sysctl -w
> > net.ipv6.conf.default.router_solicitations=1], [0], [dnl
> > +net.ipv6.conf.default.router_solicitations = 1
> > +])
> > +ADD_VETH(bob1, bob1, br-int, "172.16.0.1/16", "f0:00:00:01:02:06", \
> > + "172.16.0.2")
> > +NS_CHECK_EXEC([bob1], [ip -6 addr add fd07::1/64 dev bob1 nodad])
> > +NS_CHECK_EXEC([bob1], [ip -6 route add default via fd07::2])
> > +NS_CHECK_EXEC([bob1], [ip -6 neigh add fd07::2 lladdr 00:00:02:01:02:03
> > dev bob1])
> > +NS_CHECK_EXEC([bob1], [ip -6 neigh add fd07::3 lladdr 00:00:01:01:02:04
> > dev bob1])
> > +
> > +# Add neighbour MAC addresses to avoid sending IPv6 NS messages which could
> > +# cause datapath flows to be evicted
> > +ovn-nbctl lsp-add bob bob1 \
> > +-- lsp-set-addresses bob1 "f0:00:00:01:02:06 172.16.0.1 fd07::1"
> > +
> > +# Ensure ovn-controller is caught up
> > +ovn-nbctl --wait=hv sync
> > +
> > +on_exit 'ovs-ofctl dump-flows br-int'
> > +
> > +NETNS_DAEMONIZE([alice1], [nc -l -k 80], [alice1.pid])
> > +NS_CHECK_EXEC([bob1], [nc -z 10.0.0.2 80], [0])
> > +NS_CHECK_EXEC([bob1], [ping -q -c 3 -i 0.3 -w 2 10.0.0.2 | FORMAT_PING], \
> > +[0], [dnl
> > +3 packets transmitted, 3 received, 0% packet loss, time 0ms
> > +])
> > +
> > +# Ensure conntrack entry is present. We should not try to predict
> > +# the tunnel key for the output port, so we strip it from the labels
> > +# and just ensure that the known ethernet address is present.
> > +AT_CHECK([ovs-appctl dpctl/dump-conntrack | FORMAT_CT(172.16.0.1) | \
> > +sed -e 's/zone=[[0-9]]*/zone=<cleared>/' |
> > +sed -e 's/mark=[[0-9]]*/mark=<cleared>/' |
> > +sed -e 's/labels=0x[[0-9]]/labels=0x?/'], [0], [dnl
> > +icmp,orig=(src=172.16.0.1,dst=10.0.0.2,id=<cleared>,type=8,code=0),reply=(src=10.0.0.2,dst=172.16.0.1,id=<cleared>,type=0,code=0),zone=<cleared>,mark=<cleared>,labels=0x?000000000401020400000000
> > +tcp,orig=(src=172.16.0.1,dst=10.0.0.2,sport=<cleared>,dport=<cleared>),reply=(src=10.0.0.2,dst=172.16.0.1,sport=<cleared>,dport=<cleared>),zone=<cleared>,mark=<cleared>,labels=0x?000000000401020400000000,protoinfo=(state=<cleared>)
> > +])
> > +
> > +# Ensure datapaths show conntrack states as expected
> > +# Like with conntrack entries, we shouldn't try to predict
> > +# port binding tunnel keys. So omit them from expected labels.
> > +AT_CHECK([ovs-appctl dpctl/dump-flows | sed -e
> > 's/label=0x[[0-9]]/label=0x?/' | \
> > +grep
> > 'ct_state(+new-est-rpl+trk).*ct(.*label=0x?000000000401020400000000/.*)'
> > -c], [0], [dnl
> > +2
> > +])
> > +AT_CHECK([[ovs-appctl dpctl/dump-flows | sed -e
> > 's/ct_label(0x[0-9]/ct_label(0x?/' | \
> > +grep 'ct_state(-new+est+rpl+trk).*ct_label(0x?000000000401020400000000)'
> > -c]], [0], [dnl
> > +2
> > +])
> > +
> > +# Flush conntrack entries for easier output parsing of next test.
> > +AT_CHECK([ovs-appctl dpctl/flush-conntrack])
> > +# Change bob1 L2 address anche check the reply is properly updated.
> > +ovn-nbctl set Logical_Router_Port R2_ext mac='"00:00:10:01:02:04"'
> > +ovn-nbctl set Logical_Switch_Port r2-ext \
> > + type=router options:router-port=R2_ext addresses='"00:00:10:01:02:04"'
> > +
> > +# Wait for ovn-controller before sending traffic
> > +ovn-nbctl --wait=hv sync
> > +
> > +NS_CHECK_EXEC([bob1], [nc -z 10.0.0.2 80], [0])
> > +NS_CHECK_EXEC([bob1], [ping -q -c 3 -i 0.3 -w 2 10.0.0.2 | FORMAT_PING], \
> > +[0], [dnl
> > +3 packets transmitted, 3 received, 0% packet loss, time 0ms
> > +])
> > +AT_CHECK([ovs-appctl dpctl/dump-flows | sed -e
> > 's/label=0x[[0-9]]/label=0x?/' | \
> > +grep
> > 'ct_state(+new-est-rpl+trk).*ct(.*label=0x?000000001001020400000000/.*)'
> > -c], [0], [dnl
> > +2
> > +])
> > +AT_CHECK([[ovs-appctl dpctl/dump-flows | sed -e
> > 's/ct_label(0x[0-9]/ct_label(0x?/' | \
> > +grep 'ct_state(-new+est+rpl+trk).*ct_label(0x?000000001001020400000000)'
> > -c]], [0], [dnl
> > +2
> > +])
> > +
> > +AT_CHECK([ovs-appctl dpctl/dump-conntrack | grep 1001020400000000 |
> > FORMAT_CT(172.16.0.1) | \
> > +sed -e 's/zone=[[0-9]]*/zone=<cleared>/' |
> > +sed -e 's/mark=[[0-9]]*/mark=<cleared>/' |
> > +sed -e 's/labels=0x[[0-9]]/labels=0x?/' | sort], [0], [dnl
> > +icmp,orig=(src=172.16.0.1,dst=10.0.0.2,id=<cleared>,type=8,code=0),reply=(src=10.0.0.2,dst=172.16.0.1,id=<cleared>,type=0,code=0),zone=<cleared>,mark=<cleared>,labels=0x?000000001001020400000000
> > +tcp,orig=(src=172.16.0.1,dst=10.0.0.2,sport=<cleared>,dport=<cleared>),reply=(src=10.0.0.2,dst=172.16.0.1,sport=<cleared>,dport=<cleared>),zone=<cleared>,mark=<cleared>,labels=0x?000000001001020400000000,protoinfo=(state=<cleared>)
> > +])
> > +# Check entries in table 76 and 77 expires w/o traffic
> > +OVS_WAIT_UNTIL([
> > +test $(ovs-ofctl dump-flows br-int | grep -c 'table=OFTABLE_ECMP_NH_MAC,
> > n_packets') -eq 0
> > +])
> > +OVS_WAIT_UNTIL([
> > +test $(ovs-ofctl dump-flows br-int | grep -c 'table=OFTABLE_ECMP_NH,
> > n_packets') -eq 0
> > +])
> > +
> > +# Flush connection tracking entries
> > +ovn-nbctl --wait=hv lr-route-del R1
> > +AT_CHECK([ovs-appctl dpctl/dump-conntrack | FORMAT_CT(172.16.0.1)])
> > +
> > +# Install ECMP routes for alice.
> > +ovn-nbctl --ecmp-symmetric-reply --policy="src-ip" lr-route-add R1
> > fd01::/126 fd02::2
> > +ovn-nbctl --ecmp-symmetric-reply --policy="src-ip" lr-route-add R1
> > fd01::/126 fd02::3
> > +
> > +# Static Routes
> > +ovn-nbctl lr-route-add R2 fd01::/64 fd02::1
> > +ovn-nbctl lr-route-add R3 fd01::/64 fd02::1
> > +
> > +NETNS_DAEMONIZE([alice1], [nc -6 -l -k 8080], [alice2.pid])
> > +NS_CHECK_EXEC([bob1], [nc -6 -z fd01::2 8080], [0])
> > +NS_CHECK_EXEC([bob1], [ping -q -c 3 -i 0.3 -w 2 fd01::2 | FORMAT_PING], \
> > +[0], [dnl
> > +3 packets transmitted, 3 received, 0% packet loss, time 0ms
> > +])
> > +
> > +# Ensure conntrack entry is present. We should not try to predict
> > +# the tunnel key for the output port, so we strip it from the labels
> > +# and just ensure that the known ethernet address is present.
> > +AT_CHECK([ovs-appctl dpctl/dump-conntrack | FORMAT_CT(fd01::2) | \
> > +sed -e 's/zone=[[0-9]]*/zone=<cleared>/' |
> > +sed -e 's/mark=[[0-9]]*/mark=<cleared>/' |
> > +sed -e 's/labels=0x[[0-9]]/labels=0x?/' | sort], [0], [dnl
> > +icmpv6,orig=(src=fd07::1,dst=fd01::2,id=<cleared>,type=128,code=0),reply=(src=fd01::2,dst=fd07::1,id=<cleared>,type=129,code=0),zone=<cleared>,mark=<cleared>,labels=0x?000000001001020400000000
> > +tcp,orig=(src=fd07::1,dst=fd01::2,sport=<cleared>,dport=<cleared>),reply=(src=fd01::2,dst=fd07::1,sport=<cleared>,dport=<cleared>),zone=<cleared>,mark=<cleared>,labels=0x?000000001001020400000000,protoinfo=(state=<cleared>)
> > +])
> > +
> > +# Flush conntrack entries for easier output parsing of next test.
> > +AT_CHECK([ovs-appctl dpctl/flush-conntrack])
> > +
> > +# Change bob1 L2 address anche check the reply is properly updated.
> > +ovn-nbctl set Logical_Router_Port R2_ext mac='"00:00:10:01:02:04"'
> > +ovn-nbctl --wait=hv set Logical_Switch_Port r2-ext \
> > + type=router options:router-port=R2_ext addresses='"00:00:10:01:02:04"'
> > +
> > +NS_CHECK_EXEC([bob1], [nc -6 -z fd01::2 8080], [0])
> > +NS_CHECK_EXEC([bob1], [ping -q -c 3 -i 0.3 -w 2 fd01::2 | FORMAT_PING], \
> > +[0], [dnl
> > +3 packets transmitted, 3 received, 0% packet loss, time 0ms
> > +])
> > +
> > +AT_CHECK([ovs-appctl dpctl/dump-conntrack | grep 1001020400000000 |
> > FORMAT_CT(fd01::2) | \
> > +sed -e 's/zone=[[0-9]]*/zone=<cleared>/' |
> > +sed -e 's/mark=[[0-9]]*/mark=<cleared>/' |
> > +sed -e 's/labels=0x[[0-9]]/labels=0x?/'], [0], [dnl
> > +icmpv6,orig=(src=fd07::1,dst=fd01::2,id=<cleared>,type=128,code=0),reply=(src=fd01::2,dst=fd07::1,id=<cleared>,type=129,code=0),zone=<cleared>,mark=<cleared>,labels=0x?000000001001020400000000
> > +tcp,orig=(src=fd07::1,dst=fd01::2,sport=<cleared>,dport=<cleared>),reply=(src=fd01::2,dst=fd07::1,sport=<cleared>,dport=<cleared>),zone=<cleared>,mark=<cleared>,labels=0x?000000001001020400000000,protoinfo=(state=<cleared>)
> > +])
> > +
> > +# Flush connection tracking entries
> > +ovn-nbctl --wait=hv lr-route-del R1
> > +AT_CHECK([ovs-appctl dpctl/dump-conntrack | FORMAT_CT(fd01::2)])
> > +
> > +ovs-ofctl dump-flows br-int
> > +
> > +OVS_APP_EXIT_AND_WAIT([ovn-controller])
> > +
> > +as ovn-sb
> > +OVS_APP_EXIT_AND_WAIT([ovsdb-server])
> > +
> > +as ovn-nb
> > +OVS_APP_EXIT_AND_WAIT([ovsdb-server])
> > +
> > +as northd
> > +OVS_APP_EXIT_AND_WAIT([ovn-northd])
> > +
> > +as
> > +OVS_TRAFFIC_VSWITCHD_STOP(["/failed to query port patch-.*/d
> > +/connection dropped.*/d"])
> > +
> > +AT_CLEANUP
> > +])
> > diff --git a/tests/system-ovn.at b/tests/system-ovn.at
> > index 7ae54113a..a0f375141 100644
> > --- a/tests/system-ovn.at
> > +++ b/tests/system-ovn.at
> > @@ -6178,6 +6178,10 @@ OVS_WAIT_UNTIL([
> > test $(ovs-ofctl dump-flows br-int | grep -c 'table=OFTABLE_ECMP_NH,
> > n_packets') -eq 0
> > ])
> > +# Flush connection tracking entries
> > +ovn-nbctl --wait=hv lr-route-del R1
> > +AT_CHECK([ovs-appctl dpctl/dump-conntrack | FORMAT_CT(172.16.0.1)])
> > +
> > ovs-ofctl dump-flows br-int
> > OVS_APP_EXIT_AND_WAIT([ovn-controller])
>
_______________________________________________
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
