On 5/22/25 10:40 AM, Vasyl Saienko wrote:
> With this patch ovn-controller-vtep start synchronizing local
> MAC database to FDB OVN SB table.
> This information is used by ovn-controller instances to
> create direct flows to remove macs.
> 
> Signed-Off-By: Vasyl Saienko <vsaie...@mirantis.com>
> ---

Hi Vasyl,

Again, sorry for the long time it took to review this patch.

I have a few comments.  Most of them are nits that I could fix up myself
but the main question is related to SB.FDB entries created by
ovn-controller-vtep and the fact that if we have FDB aging enabled
things might behave a bit weird as ovn-northd would continuously remove
(age-out) these entries.  Please see inline.

Also, with this patch applied, the "vtep: 3 HVs, 1 VIFs/HV, 1 GW, 1 LS"
test becomes unstable because there's a race condition as the test is
not waiting for SB.FDB entries to be processed by ovn-controller.
One way to avoid that could be to add the following incremental change:

diff --git a/tests/ovn.at b/tests/ovn.at
index 108de260ef..b0cc95d224 100644
--- a/tests/ovn.at
+++ b/tests/ovn.at
@@ -4678,6 +4678,12 @@ AT_CHECK([ovn-sbctl lflow-list lsw0 | grep 
'reg0[\[14\]]' | ovn_strip_lflows], [
   table=??(ls_in_hairpin      ), priority=2000 , match=(reg0[[14]] == 1 && 
is_chassis_resident("cr-lrp2")), action=(next;)
 ])
 
+# Ensure FDB table is up to date.
+wait_row_count FDB 1 mac='"02:00:00:00:00:ee"'
+wait_row_count FDB 1 mac='"02:00:00:00:00:ff"'
+wait_row_count FDB 1 mac='"f0:00:00:00:00:03"'
+check_row_count FDB 3
+
 # dump information with counters
 echo "------ OVN dump ------"
 ovn-nbctl show

The test failed a couple of times in CI:
https://github.com/ovsrobot/ovn/actions/runs/15183730341/job/42702264742#step:11:5010

>  controller-vtep/ovn-controller-vtep.c |   6 +
>  controller-vtep/vtep.c                | 173 ++++++++++++++++++++++++++
>  tests/ovn-controller-vtep.at          | 102 ++++++++++++++-
>  3 files changed, 279 insertions(+), 2 deletions(-)
> 
> diff --git a/controller-vtep/ovn-controller-vtep.c 
> b/controller-vtep/ovn-controller-vtep.c
> index 698511482..ea113f8df 100644
> --- a/controller-vtep/ovn-controller-vtep.c
> +++ b/controller-vtep/ovn-controller-vtep.c
> @@ -171,6 +171,12 @@ main(int argc, char *argv[])
>      ovsdb_idl_add_column(ovnsb_idl_loop.idl, 
> &sbrec_port_binding_col_options);
>      ovsdb_idl_add_column(ovnsb_idl_loop.idl, &sbrec_port_binding_col_type);
>      ovsdb_idl_add_column(ovnsb_idl_loop.idl, &sbrec_port_binding_col_up);
> +    ovsdb_idl_add_column(ovnsb_idl_loop.idl,
> +                         &sbrec_port_binding_col_tunnel_key);
> +    ovsdb_idl_add_table(ovnsb_idl_loop.idl, &sbrec_table_fdb);
> +    ovsdb_idl_add_column(ovnsb_idl_loop.idl, &sbrec_fdb_col_mac);
> +    ovsdb_idl_add_column(ovnsb_idl_loop.idl, &sbrec_fdb_col_dp_key);
> +    ovsdb_idl_add_column(ovnsb_idl_loop.idl, &sbrec_fdb_col_port_key);
>  
>      ovsdb_idl_set_leader_only(ovnsb_idl_loop.idl, false);
>      ovsdb_idl_get_initial_snapshot(ovnsb_idl_loop.idl);
> diff --git a/controller-vtep/vtep.c b/controller-vtep/vtep.c
> index 7f5e1d606..f86cb679f 100644
> --- a/controller-vtep/vtep.c
> +++ b/controller-vtep/vtep.c
> @@ -42,6 +42,12 @@ struct mmr_hash_node_data {
>      struct shash physical_locators;
>  };
>  
> +struct lp_fdb_node_data {
> +    uint32_t dp_key;
> +    uint32_t port_key;
> +    struct shash mac_fdbs;
> +};
> +
>  /*
>   * Scans through the Binding table in ovnsb, and updates the vtep logical
>   * switch tunnel keys and the 'Ucast_Macs_Remote' table in the VTEP
> @@ -519,6 +525,140 @@ vtep_mcast_macs_cleanup(struct ovsdb_idl *vtep_idl)
>  
>      return true;
>  }
> +
> +static const struct sbrec_port_binding *
> +find_pbs_for_logical_switch(struct shash *vtep_pbs,

Nit: I'd call this logical_switch_find_pb().

> +                            struct sset *vtep_pswitches,
> +                            char *ls_name){
> +
> +    struct shash_node *node;
> +    SHASH_FOR_EACH (node, vtep_pbs) {
> +        const struct sbrec_port_binding *port_binding_rec = node->data;
> +        const char *pswitch_name = smap_get(&port_binding_rec->options,
> +                                            "vtep-physical-switch");
> +        const char *lswitch_name = smap_get(&port_binding_rec->options,
> +                                            "vtep-logical-switch");
> +        if (!port_binding_rec->chassis) {
> +            continue;
> +        }
> +
> +        /* If 'port_binding_rec->chassis' exists then 'pswitch_name'
> +         * and 'lswitch_name' must also exist. */
> +        if (!pswitch_name || !lswitch_name) {
> +            /* This could only happen when someone directly modifies the
> +             * database,  (e.g. using ovn-sbctl). */
> +            VLOG_ERR("logical port (%s) with no 'options:vtep-physical-"

Nit: VLOG_ERR_RL() instead.

> +                     "switch' or 'options:vtep-logical-switch' specified "
> +                     "is bound to chassis (%s).",
> +                     port_binding_rec->logical_port,
> +                     port_binding_rec->chassis->name);
> +            continue;
> +        }
> +        /* Make sure both logical_switch and physical_switch matches */
> +        if (!strcmp(ls_name, lswitch_name)) {
> +          if (sset_find(vtep_pswitches, port_binding_rec->chassis->name)) {
> +            return port_binding_rec;
> +          }
> +        }

Nit: indentation is off here.

> +    }
> +    return NULL;
> +}
> +
> +/* Propagate dynamically learned MACs on local VTEPs to OVN SB. Where
> + * later this information is used to create tunnels between neighbour
> + * VTEPs.
> +*/
> +static void
> +vtep_local_macs(struct controller_vtep_ctx *ctx,
> +                struct shash *vtep_pbs,
> +                struct sset *vtep_pswitches,
> +                struct shash *vtep_lswitches,
> +                struct shash *lp_fdb){
> +
> +    if (!ctx->ovnsb_idl_txn) {
> +        return;
> +    }
> +
> +    ovsdb_idl_txn_add_comment(ctx->ovnsb_idl_txn,
> +                              "ovn-controller-vtep: updating fdb");
> +
> +    const struct vteprec_ucast_macs_local *vtep_uml;
> +    const struct sbrec_port_binding *port_binding_rec;
> +    const struct sbrec_fdb *fdb;

These last two can be declared inside the loops, restricting the scope of the 
variables.

> +
> +    /* Collect local unicast MACs */
> +    VTEPREC_UCAST_MACS_LOCAL_FOR_EACH (vtep_uml, ctx->vtep_idl) {
> +        port_binding_rec = find_pbs_for_logical_switch(
> +            vtep_pbs, vtep_pswitches, vtep_uml->logical_switch->name);
> +        if (!port_binding_rec) {
> +            VLOG_ERR("Cannot find port_binding for dynamically learned MAC "
> +                     "%s in logical_switch %s", vtep_uml->MAC,
> +                     vtep_uml->logical_switch->name);
> +            continue;
> +        }
> +
> +        if (!port_binding_rec->datapath->tunnel_key ||
> +            !port_binding_rec->tunnel_key) {
> +            continue;
> +        }

Datapath_bindings and port binding tunnel_keys can never be 0.  We can skip 
this check.

> +
> +        char *fdb_dp_port_key = xasprintf(
> +            "%"PRId64"_%"PRId64, port_binding_rec->datapath->tunnel_key,
> +            port_binding_rec->tunnel_key);
> +
> +        struct lp_fdb_node_data *sbrec_lp_fdb = shash_find_data(
> +            lp_fdb, fdb_dp_port_key);

The indentation looks a bit awkward here.  I'd change it to:

        struct lp_fdb_node_data *sbrec_lp_fdb =
            shash_find_data(lp_fdb, fdb_dp_port_key);

> +        fdb = NULL;
> +        if (sbrec_lp_fdb) {
> +            fdb = shash_find_and_delete(&sbrec_lp_fdb->mac_fdbs,
> +                                        vtep_uml->MAC);
> +        }
> +        free(fdb_dp_port_key);
> +
> +        if (!fdb) {
> +            const struct sbrec_fdb *sb_fdb;
> +            VLOG_DBG("Creating new FDB entry for mac %s", vtep_uml->MAC);
> +            sb_fdb = sbrec_fdb_insert(ctx->ovnsb_idl_txn);
> +            sbrec_fdb_set_dp_key(sb_fdb,
> +                                 port_binding_rec->datapath->tunnel_key);
> +            sbrec_fdb_set_mac(sb_fdb, vtep_uml->MAC);
> +            sbrec_fdb_set_port_key(sb_fdb, port_binding_rec->tunnel_key);
> +            sbrec_fdb_set_timestamp(sb_fdb, time_wall_msec());

So, this makes me wonder, how does this new feature work in conjunction with
FDB aging (ovn-northd removes FDB entries that are older and inactive for more
than a configured threshold).  I'm guess it doesn't work.
Do we need to support that case?

If yes, do we need to periodically refresh the timestamp of vtep-inserted
FDB records?

> +        }
> +    }
> +
> +    struct shash_node *node;
> +    struct shash_node *sbrec_fdb_node;

This can be declared inside the loop, restricting the scope of the variable.

> +    SHASH_FOR_EACH (node, vtep_lswitches) {
> +        port_binding_rec = find_pbs_for_logical_switch(vtep_pbs,
> +                                                       vtep_pswitches,
> +                                                       node->name);
> +        if (!port_binding_rec) {
> +            continue;
> +        }
> +        if (!port_binding_rec->datapath->tunnel_key ||
> +            !port_binding_rec->tunnel_key) {
> +                continue;
> +        }
> +
> +        char *fdb_dp_port_key = xasprintf(
> +            "%"PRId64"_%"PRId64, port_binding_rec->datapath->tunnel_key,
> +            port_binding_rec->tunnel_key);
> +        struct lp_fdb_node_data *sbrec_lp_fdb = shash_find_data(
> +            lp_fdb, fdb_dp_port_key);
> +        free(fdb_dp_port_key);
> +
> +        if (!sbrec_lp_fdb) {
> +            continue;
> +        }
> +
> +        SHASH_FOR_EACH (sbrec_fdb_node, &sbrec_lp_fdb->mac_fdbs) {
> +            struct sbrec_fdb *mac_fdb = sbrec_fdb_node->data;
> +            VLOG_DBG("Removing stale FDB for VTEP mac %s", mac_fdb->mac);
> +            sbrec_fdb_delete(mac_fdb);
> +        }
> +    }
> +}
>  
>  /* Updates vtep logical switch tunnel keys. */
>  void
> @@ -605,6 +745,29 @@ vtep_run(struct controller_vtep_ctx *ctx)
>          shash_add(target, port_binding_rec->logical_port, port_binding_rec);
>      }
>  
> +    /* Construct logical_port to fdb */
> +    const struct sbrec_fdb *fdb;
> +    struct shash lp_fdbs = SHASH_INITIALIZER(&lp_fdbs);
> +
> +    SBREC_FDB_FOR_EACH (fdb, ctx->ovnsb_idl) {
> +        char *fdb_dp_port_key = xasprintf("%"PRId64"_%"PRId64, fdb->dp_key,
> +                                          fdb->port_key);
> +        struct lp_fdb_node_data *lp_fdb = shash_find_data(&lp_fdbs,
> +                                                          fdb_dp_port_key);
> +

Nit: we don't really need this empty line here.

> +        if (!lp_fdb) {
> +            lp_fdb = xmalloc(sizeof *lp_fdb);
> +            lp_fdb->dp_key = fdb->dp_key;
> +            lp_fdb->port_key = fdb->port_key;
> +            shash_init(&lp_fdb->mac_fdbs);

To ensure that all members of the structure are initialized (at least zeroed),
in the eventuality that new fields are added in the future, I'd write this as:

            lp_fdb = xmalloc(sizeof *lp_fdb);
            *lp_fdb = (struct lp_fdb_node_data) {
                .dp_key = fdb->dp_key,
                .port_key = fdb->port_key,
                .mac_fdbs = SHASH_INITIALIZER(&lp_fdb->mac_fdbs),
            };

Nit: I'd add an empty line here between member initializations and other
operations.

> +            shash_add(&lp_fdb->mac_fdbs, fdb->mac, fdb);

This is executed in both cases (lp_fdb existed or not), it can be moved after
the "if/else".

> +            shash_add(&lp_fdbs, fdb_dp_port_key, lp_fdb);
> +        } else {
> +            shash_add(&lp_fdb->mac_fdbs, fdb->mac, fdb);
> +        }
> +        free(fdb_dp_port_key);
> +    }
> +
>      ovsdb_idl_txn_add_comment(ctx->vtep_idl_txn,
>                                "ovn-controller-vtep: update logical switch "
>                                "tunnel keys and 'ucast_macs_remote's");
> @@ -613,6 +776,8 @@ vtep_run(struct controller_vtep_ctx *ctx)
>      vtep_macs_run(ctx->vtep_idl_txn, &ucast_macs_rmts,
>                    &mcast_macs_rmts, &physical_locators,
>                    &vtep_lswitches, &non_vtep_pbs);
> +    vtep_local_macs(ctx, &vtep_pbs, &vtep_pswitches, &vtep_lswitches,
> +                    &lp_fdbs);
>  
>      sset_destroy(&vtep_pswitches);
>      shash_destroy(&vtep_lswitches);
> @@ -628,6 +793,14 @@ vtep_run(struct controller_vtep_ctx *ctx)
>      shash_destroy(&physical_locators);
>      shash_destroy(&vtep_pbs);
>      shash_destroy(&non_vtep_pbs);
> +
> +    struct shash_node *lp_fdb_node;
> +    SHASH_FOR_EACH (lp_fdb_node, &lp_fdbs) {
> +        struct lp_fdb_node_data *lp_fdb = lp_fdb_node->data;
> +        shash_destroy(&lp_fdb->mac_fdbs);
> +        free(lp_fdb);
> +    }
> +    shash_destroy(&lp_fdbs);
>  }
>  
>  /* Cleans up all related entries in vtep.  Returns true when done (i.e. there
> diff --git a/tests/ovn-controller-vtep.at b/tests/ovn-controller-vtep.at
> index d410ecd28..2434d9126 100644
> --- a/tests/ovn-controller-vtep.at
> +++ b/tests/ovn-controller-vtep.at
> @@ -1,9 +1,10 @@
>  AT_BANNER([ovn_controller_vtep])
>  
> -# OVN_CONTROLLER_VTEP_START(SIM_NAME)
> +# OVN_CONTROLLER_VTEP_START(SIM_NAME, TUNNEL_IP)
>  #
>  # $1 - optional simulator name. If none is given, runs ovn-controller-vtep, 
> and
>  #      vtep emulator in $ovs_dir.
> +# $2 - optional tunnel ip. If none is given default 1.2.3.4 is used
>  # Starts the test with a setup with vtep device.  Each test case must first
>  # call this macro and ovn_start.
>  #
> @@ -14,7 +15,9 @@ m4_define([OVN_CONTROLLER_VTEP_START], [
>     AT_KEYWORDS([ovn])
>     # this will cause skip when 'make check' using Windows setup.
>     sim="$1"
> +   tunnel_ip="$2"
>     vtep_chassis=${sim:-br-vtep}
> +   tunnel_ip=${tunnel_ip:-1.2.3.4}
>  
>     test -n "$sim" && as "$sim"
>     mkdir -p "$ovs_dir" || return 1
> @@ -41,7 +44,7 @@ m4_define([OVN_CONTROLLER_VTEP_START], [
>                -- add-port $vtep_chassis p1 -- set Interface p1 type=dummy 
> ofport_request=2
>  
>     dnl Start ovs-vtep.
> -   check vtep-ctl add-ps $vtep_chassis -- set Physical_Switch $vtep_chassis 
> tunnel_ips=1.2.3.4
> +   check vtep-ctl add-ps $vtep_chassis -- set Physical_Switch $vtep_chassis 
> tunnel_ips=$tunnel_ip
>     AT_CHECK([ovs-vtep --log-file="$ovs_dir"/ovs-vtep.log \
>                        --pidfile="$ovs_dir"/ovs-vtep.pid \
>                        --detach --no-chdir $vtep_chassis], [0], [], [stderr])
> @@ -86,6 +89,9 @@ AT_CHECK([ovn-nbctl lsp-set-type $2 vtep])
>  AT_CHECK([ovn-nbctl lsp-set-options $2 vtep-physical-switch=$3 
> vtep-logical-switch=$4])
>  ])
>  
> +# mac learning table is hardcoded in ovs-vtep emulator
> +m4_define([mac_learning_table], [1])
> +
>  ##############################################
>  
>  # tests chassis related updates.
> @@ -509,6 +515,98 @@ AT_CHECK([vtep-ctl --columns=MAC list Ucast_Macs_Remote 
> | cut -d ':' -f2- | tr -
>  OVN_CONTROLLER_VTEP_STOP([/has already been known to be on logical port/d])
>  AT_CLEANUP
>  
> +# Tests vtep module 'Ucast_Macs_Local' MACs propagation
> +AT_SETUP([ovn-controller-vtep - vtep-macs 3])
> +ovn_start
> +OVN_CONTROLLER_VTEP_START
> +
> +check ovn-nbctl ls-add br-test
> +check ovn-nbctl ls-add br-test1
> +
> +# creates a simple logical network with the vtep device and a fake hv chassis
> +# 'ch0'.
> +AT_CHECK([ovn-nbctl --wait=sb sync])
> +
> +# creates the logical switch in vtep and adds the corresponding logical
> +# port to 'br-test'.
> +AT_CHECK([vtep-ctl add-ls lswitch0 -- bind-ls br-vtep p0 100 lswitch0])
> +OVN_NB_ADD_VTEP_PORT([br-test], [br-vtep_lswitch0], [br-vtep], [lswitch0])
> +OVS_WAIT_UNTIL([test -n "`ovn-sbctl list Port_Binding  | grep 
> br-vtep_lswitch0`"])

Nit: please use wait_row_count, e.g.:

wait_row_count Port_Binding 1 logical_port=br-vtep_lswitch0

> +OVS_WAIT_UNTIL([test -n "`ovs-vsctl show | grep br-vtep_vtep_ls1`"])
> +
> +# Simulate peer on physical port to populate Ucast_Macs_Local table
> +AT_CHECK([ovs-ofctl -O OpenFlow14 add-flow br-vtep_vtep_ls1 
> "cookie=0x5000,table=mac_learning_table,priority=1000,dl_dst=aa:bb:cc:dd:ee:01
>  actions=output:0100-p0-l"])
> +
> +# Wait Ucast_Macs_Local is updated with local MAC
> +OVS_WAIT_UNTIL([test -n "`vtep-ctl list Ucast_Macs_Local | grep 
> aa:bb:cc:dd:ee:01`"])
> +
> +# Check that OVNSB is updated with learned MAC
> +OVS_WAIT_UNTIL([test -n "`ovn-sbctl find Fdb mac='"aa:bb:cc:dd:ee:01"' |grep 
> _uuid`"])
> +
> +# Plug p1 to lswitch1
> +AT_CHECK([vtep-ctl add-ls lswitch1 -- bind-ls br-vtep p1 100 lswitch1])
> +OVN_NB_ADD_VTEP_PORT([br-test1], [br-vtep_lswitch1], [br-vtep], [lswitch1])
> +OVS_WAIT_UNTIL([test -n "`ovn-sbctl list Port_Binding  | grep 
> br-vtep_lswitch1`"])

Nit: wait_row_count Port_Binding 1 logical_port=br-vtep_lswitch1

> +OVS_WAIT_UNTIL([test -n "`ovs-vsctl show | grep br-vtep_vtep_ls2`"])
> +
> +OVS_WAIT_UNTIL([test -n "`ovs-vsctl show |grep 0100-p1-l`"])
> +
> +# Simulate peer on physical port to populate Ucast_Macs_Local table
> +AT_CHECK([ovs-ofctl -O OpenFlow14 add-flow br-vtep_vtep_ls2 
> "cookie=0x5000,table=mac_learning_table,priority=1000,dl_dst=aa:bb:cc:dd:ee:02
>  actions=output:0100-p1-l"])
> +
> +# Wait Ucast_Macs_Local is updated with local MAC
> +OVS_WAIT_UNTIL([test -n "`vtep-ctl list Ucast_Macs_Local | grep 
> aa:bb:cc:dd:ee:02`"])
> +
> +# Remove MAC from Ucast_Macs_Local
> +AT_CHECK([ovs-ofctl -O OpenFlow14 del-flows br-vtep_vtep_ls1 
> "table=mac_learning_table,dl_dst=aa:bb:cc:dd:ee:01"])
> +
> +# Check that first MAC is removed from OVNSB, but second is still there
> +OVS_WAIT_UNTIL([test -z "`ovn-sbctl find Fdb mac='"aa:bb:cc:dd:ee:01"'`"])
> +OVS_WAIT_UNTIL([test -n "`ovn-sbctl find Fdb mac='"aa:bb:cc:dd:ee:02"'`"])
> +

Nit: wait_row_count

> +# Remove second MAC from Ucast_Macs_Local
> +AT_CHECK([ovs-ofctl -O OpenFlow14 del-flows br-vtep_vtep_ls2 
> "table=mac_learning_table,dl_dst=aa:bb:cc:dd:ee:02"])
> +
> +# Check that second MAC is removed from OVNSB
> +OVS_WAIT_UNTIL([test -z "`ovn-sbctl find Fdb mac='"aa:bb:cc:dd:ee:01"'`"])
> +OVS_WAIT_UNTIL([test -z "`ovn-sbctl find Fdb mac='"aa:bb:cc:dd:ee:02"'`"])
> +
> +# Start second VTEP chassis
> +OVN_CONTROLLER_VTEP_START(br-vtep1, 1.2.3.5)
> +
> +# creates the logical switch in vtep and adds the corresponding logical
> +# # port to 'br-test'.
> +AT_CHECK([vtep-ctl add-ls lswitch0 -- bind-ls br-vtep1 p0 100 lswitch0])
> +OVN_NB_ADD_VTEP_PORT([br-test], [br-vtep1_lswitch0], [br-vtep1], [lswitch0])
> +OVS_WAIT_UNTIL([test -n "`ovn-sbctl list Port_Binding  | grep 
> br-vtep1_lswitch0`"])

Nit: wait_row_count

> +OVS_WAIT_UNTIL([test -n "`ovs-vsctl show | grep br-vtep1_vtep_ls1`"])
> +
> +# creates the logical switch in vtep and adds the corresponding logical
> +# # # port to 'br-test'.
> +AT_CHECK([vtep-ctl add-ls lswitch1 -- bind-ls br-vtep1 p1 100 lswitch1])
> +OVN_NB_ADD_VTEP_PORT([br-test1], [br-vtep1_lswitch1], [br-vtep1], [lswitch1])
> +OVS_WAIT_UNTIL([test -n "`ovn-sbctl list Port_Binding  | grep 
> br-vtep1_lswitch1`"])

Nit: wait_row_count

> +OVS_WAIT_UNTIL([test -n "`ovs-vsctl show | grep br-vtep1_vtep_ls2`"])
> +
> +# Wait bingins are activated
> +OVS_WAIT_UNTIL([test -n "`ovs-vsctl show |grep 0100-p0-l`"])
> +OVS_WAIT_UNTIL([test -n "`ovs-vsctl show |grep 0100-p1-l`"])
> +
> +# Simulate peer on physical port to populate Ucast_Macs_Local table on 
> br-vtep1
> +AT_CHECK([ovs-ofctl -O OpenFlow14 add-flow br-vtep1_vtep_ls1 
> "cookie=0x5000,table=mac_learning_table,priority=1000,dl_dst=aa:bb:cc:dd:ff:01
>  actions=output:0100-p0-l"])
> +AT_CHECK([ovs-ofctl -O OpenFlow14 add-flow br-vtep1_vtep_ls2 
> "cookie=0x5000,table=mac_learning_table,priority=1000,dl_dst=aa:bb:cc:dd:ff:02
>  actions=output:0100-p1-l"])
> +
> +# Wait Ucast_Macs_Local is updated with local MAC
> +OVS_WAIT_UNTIL([test -n "`vtep-ctl list Ucast_Macs_Local | grep 
> aa:bb:cc:dd:ff:01`"])
> +OVS_WAIT_UNTIL([test -n "`vtep-ctl list Ucast_Macs_Local | grep 
> aa:bb:cc:dd:ff:02`"])
> +
> +# Check that first MACs is propagated to FDB table
> +OVS_WAIT_UNTIL([test -n "`ovn-sbctl find Fdb mac='"aa:bb:cc:dd:ff:01"'`"])
> +OVS_WAIT_UNTIL([test -n "`ovn-sbctl find Fdb mac='"aa:bb:cc:dd:ff:02"'`"])

Nit: wait_row_count

> +
> +OVN_CONTROLLER_VTEP_STOP
> +AT_CLEANUP
> +
>  # Tests vtep module 'Mcast_Macs_Remote's.
>  AT_SETUP([ovn-controller-vtep - vtep-Mcast_Macs_Remote])
>  ovn_start

Regards,
Dumitru

_______________________________________________
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Reply via email to