On Tue, Jun 7, 2022 at 9:07 AM Numan Siddique <num...@ovn.org> wrote:
>
> On Mon, Jun 6, 2022 at 6:44 PM Lorenzo Bianconi
> <lorenzo.bianc...@redhat.com> wrote:
> >
> > Add the capability to automatically deploy a load-balancer on each
> > logical-switch connected to a logical router where the load-balancer
> > has been installed by the CMS. This patch allow to overcome the
> > distributed gw router scenario limitation where a load-balancer must be
> > installed on each datapath to properly reach the load-balancer.
> >
> > Reported-at: https://bugzilla.redhat.com/show_bug.cgi?id=2043543
> > Signed-off-by: Lorenzo Bianconi <lorenzo.bianc...@redhat.com>
>
> Thanks for the rebase.
>
> Acked-by: Numan Siddique <num...@ovn.org>
I applied this patch to the main.
Numan
>
> Numan
>
> > ---
> > Changes since v2:
> > - rebase on top of ovn master
> >
> > Changes since v1:
> > - rebase on top of ovn master
> > - add NEWS entry
> > - improve selftests
> > ---
> > NEWS | 5 +++
> > northd/northd.c | 56 +++++++++++++++++++++++++++++
> > northd/ovn-northd.8.xml | 8 +++++
> > tests/ovn-northd.at | 80 +++++++++++++++++++++++++++++++++++++++++
> > 4 files changed, 149 insertions(+)
> >
> > diff --git a/NEWS b/NEWS
> > index e015ae8e7..8b4f91553 100644
> > --- a/NEWS
> > +++ b/NEWS
> > @@ -4,6 +4,11 @@ Post v22.06.0
> > "ovn-encap-df_default" to enable or disable tunnel DF flag.
> > - Add option "localnet_learn_fdb" to LSP that will allow localnet
> > ports to learn MAC addresses and store them in FDB table.
> > + - northd: introduce the capability to automatically deploy a
> > load-balancer
> > + on each logical-switch connected to a logical router where the
> > + load-balancer has been installed by the CMS. In order to enable the
> > + feature the CMS has to set install_ls_lb_from_router to true in option
> > + column of NB_Global table.
> >
> > OVN v22.06.0 - XX XXX XXXX
> > --------------------------
> > diff --git a/northd/northd.c b/northd/northd.c
> > index 0207f6ce1..a95a5148e 100644
> > --- a/northd/northd.c
> > +++ b/northd/northd.c
> > @@ -63,6 +63,8 @@ static bool lflow_hash_lock_initialized = false;
> >
> > static bool check_lsp_is_up;
> >
> > +static bool install_ls_lb_from_router;
> > +
> > /* MAC allocated for service monitor usage. Just one mac is allocated
> > * for this purpose and ovn-controller's on each chassis will make use
> > * of this mac when sending out the packets to monitor the services
> > @@ -4140,6 +4142,55 @@ build_lrouter_lbs_reachable_ips(struct hmap
> > *datapaths, struct hmap *lbs)
> > }
> > }
> >
> > +static void
> > +build_lswitch_lbs_from_lrouter(struct hmap *datapaths, struct hmap *lbs)
> > +{
> > + if (!install_ls_lb_from_router) {
> > + return;
> > + }
> > +
> > + struct ovn_datapath *od;
> > + HMAP_FOR_EACH (od, key_node, datapaths) {
> > + if (!od->nbs) {
> > + continue;
> > + }
> > +
> > + struct ovn_port *op;
> > + LIST_FOR_EACH (op, dp_node, &od->port_list) {
> > + if (!lsp_is_router(op->nbsp)) {
> > + continue;
> > + }
> > + if (!op->peer) {
> > + continue;
> > + }
> > +
> > + struct ovn_datapath *peer_od = op->peer->od;
> > + for (size_t i = 0; i < peer_od->nbr->n_load_balancer; i++) {
> > + bool installed = false;
> > + const struct uuid *lb_uuid =
> > + &peer_od->nbr->load_balancer[i]->header_.uuid;
> > + struct ovn_northd_lb *lb = ovn_northd_lb_find(lbs,
> > lb_uuid);
> > + if (!lb) {
> > + continue;
> > + }
> > +
> > + for (size_t j = 0; j < lb->n_nb_ls; j++) {
> > + if (lb->nb_ls[j] == od) {
> > + installed = true;
> > + break;
> > + }
> > + }
> > + if (!installed) {
> > + ovn_northd_lb_add_ls(lb, od);
> > + }
> > + if (lb->nlb) {
> > + od->has_lb_vip |= lb_has_vip(lb->nlb);
> > + }
> > + }
> > + }
> > + }
> > +}
> > +
> > /* This must be called after all ports have been processed, i.e., after
> > * build_ports() because the reachability check requires the router ports
> > * networks to have been parsed.
> > @@ -4152,6 +4203,7 @@ build_lb_port_related_data(struct hmap *datapaths,
> > struct hmap *ports,
> > build_lrouter_lbs_check(datapaths);
> > build_lrouter_lbs_reachable_ips(datapaths, lbs);
> > build_lb_svcs(input_data, ovnsb_txn, ports, lbs);
> > + build_lswitch_lbs_from_lrouter(datapaths, lbs);
> > }
> >
> > /* Syncs relevant load balancers (applied to logical switches) to the
> > @@ -15378,6 +15430,10 @@ ovnnb_db_run(struct northd_input *input_data,
> > "ignore_lsp_down", true);
> > default_acl_drop = smap_get_bool(&nb->options, "default_acl_drop",
> > false);
> >
> > + install_ls_lb_from_router = smap_get_bool(&nb->options,
> > + "install_ls_lb_from_router",
> > + false);
> > +
> > build_chassis_features(input_data, &data->features);
> > build_datapaths(input_data, ovnsb_txn, &data->datapaths,
> > &data->lr_list);
> > build_lbs(input_data, &data->datapaths, &data->lbs);
> > diff --git a/northd/ovn-northd.8.xml b/northd/ovn-northd.8.xml
> > index 1f7022490..2a2b33051 100644
> > --- a/northd/ovn-northd.8.xml
> > +++ b/northd/ovn-northd.8.xml
> > @@ -882,6 +882,10 @@
> > <code>reg2</code>. For IPv6 traffic the flow also loads the
> > original
> > destination IP and transport port in registers <code>xxreg1</code>
> > and
> > <code>reg2</code>.
> > + The above flow is created even if the load balancer is attached to
> > a
> > + logical router connected to the current logical switch and
> > + the <code>install_ls_lb_from_router</code> variable in
> > + <ref table="NB_Global" column="options"/> is set to true.
> > </li>
> > <li>
> > For all the configured load balancing rules for a switch in
> > @@ -898,6 +902,10 @@
> > <code>reg2</code>. For IPv6 traffic the flow also loads the
> > original
> > destination IP and transport port in registers <code>xxreg1</code>
> > and
> > <code>reg2</code>.
> > + The above flow is created even if the load balancer is attached to
> > a
> > + logical router connected to the current logical switch and
> > + the <code>install_ls_lb_from_router</code> variable in
> > + <ref table="NB_Global" column="options"/> is set to true.
> > </li>
> >
> > <li>
> > diff --git a/tests/ovn-northd.at b/tests/ovn-northd.at
> > index a94a7d441..c1ec9e04a 100644
> > --- a/tests/ovn-northd.at
> > +++ b/tests/ovn-northd.at
> > @@ -7613,3 +7613,83 @@ AT_CHECK([ovn-sbctl dump-flows ls0 | grep -e
> > 'ls_in_\(put\|lookup\)_fdb' | sort
> >
> > AT_CLEANUP
> > ])
> > +
> > +AT_SETUP([check install_ls_lb_from_router option])
> > +AT_KEYWORDS([lb-ls-install-from-lrouter])
> > +ovn_start
> > +
> > +ovn-nbctl lr-add R1
> > +ovn-nbctl set logical_router R1 options:chassis=hv1
> > +ovn-nbctl lrp-add R1 R1-S0 02:ac:10:01:00:01 10.0.0.1/24
> > +ovn-nbctl lrp-add R1 R1-S1 02:ac:10:01:01:01 20.0.0.1/24
> > +ovn-nbctl lrp-add R1 R1-PUB 02:ac:20:01:01:01 172.16.0.1/24
> > +
> > +ovn-nbctl ls-add S0
> > +ovn-nbctl lsp-add S0 S0-R1
> > +ovn-nbctl lsp-set-type S0-R1 router
> > +ovn-nbctl lsp-set-addresses S0-R1 02:ac:10:01:00:01
> > +ovn-nbctl lsp-set-options S0-R1 router-port=R1-S0
> > +
> > +ovn-nbctl ls-add S1
> > +ovn-nbctl lsp-add S1 S1-R1
> > +ovn-nbctl lsp-set-type S1-R1 router
> > +ovn-nbctl lsp-set-addresses S1-R1 02:ac:10:01:01:01
> > +ovn-nbctl lsp-set-options S1-R1 router-port=R1-S1
> > +
> > +# Add load balancers on the logical router R1
> > +ovn-nbctl lb-add lb0 172.16.0.10:80 10.0.0.2:80
> > +ovn-nbctl lr-lb-add R1 lb0
> > +
> > +ovn-sbctl dump-flows S0 > S0flows
> > +ovn-sbctl dump-flows S1 > S1flows
> > +
> > +AT_CAPTURE_FILE([S0flows])
> > +AT_CAPTURE_FILE([S1flows])
> > +
> > +AT_CHECK([grep "ls_in_lb" S0flows | sort], [0], [dnl
> > + table=11(ls_in_lb ), priority=0 , match=(1), action=(next;)
> > +])
> > +AT_CHECK([grep "ls_in_lb" S1flows | sort], [0], [dnl
> > + table=11(ls_in_lb ), priority=0 , match=(1), action=(next;)
> > +])
> > +
> > +ovn-nbctl --wait=sb set NB_Global . options:install_ls_lb_from_router=true
> > +
> > +ovn-sbctl dump-flows S0 > S0flows
> > +ovn-sbctl dump-flows S1 > S1flows
> > +
> > +AT_CAPTURE_FILE([S0flows])
> > +AT_CAPTURE_FILE([S1flows])
> > +
> > +AT_CHECK([grep "ls_in_lb" S0flows | sort], [0], [dnl
> > + table=11(ls_in_lb ), priority=0 , match=(1), action=(next;)
> > + table=11(ls_in_lb ), priority=120 , match=(ct.new && ip4.dst
> > == 172.16.0.10 && tcp.dst == 80), action=(reg0[[1]] = 0; reg1 =
> > 172.16.0.10; reg2[[0..15]] = 80; ct_lb_mark(backends=10.0.0.2:80);)
> > +])
> > +AT_CHECK([grep "ls_in_lb" S1flows | sort], [0], [dnl
> > + table=11(ls_in_lb ), priority=0 , match=(1), action=(next;)
> > + table=11(ls_in_lb ), priority=120 , match=(ct.new && ip4.dst
> > == 172.16.0.10 && tcp.dst == 80), action=(reg0[[1]] = 0; reg1 =
> > 172.16.0.10; reg2[[0..15]] = 80; ct_lb_mark(backends=10.0.0.2:80);)
> > +])
> > +
> > +s0_uuid=$(ovn-sbctl get datapath S0 _uuid)
> > +s1_uuid=$(ovn-sbctl get datapath S1 _uuid)
> > +check_column "$s0_uuid $s1_uuid" sb:load_balancer datapaths name=lb0
> > +
> > +ovn-nbctl --wait=sb set NB_Global . options:install_ls_lb_from_router=false
> > +
> > +ovn-sbctl dump-flows S0 > S0flows
> > +ovn-sbctl dump-flows S1 > S1flows
> > +
> > +AT_CAPTURE_FILE([S0flows])
> > +AT_CAPTURE_FILE([S1flows])
> > +
> > +AT_CHECK([grep "ls_in_lb" S0flows | sort], [0], [dnl
> > + table=11(ls_in_lb ), priority=0 , match=(1), action=(next;)
> > +])
> > +AT_CHECK([grep "ls_in_lb" S1flows | sort], [0], [dnl
> > + table=11(ls_in_lb ), priority=0 , match=(1), action=(next;)
> > +])
> > +
> > +check_column "" sb:load_balancer datapaths name=lb0
> > +
> > +AT_CLEANUP
> > +])
> > --
> > 2.35.3
> >
> > _______________________________________________
> > dev mailing list
> > d...@openvswitch.org
> > https://mail.openvswitch.org/mailman/listinfo/ovs-dev
> >
_______________________________________________
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
