fix(port): change TRUNK to ACCESS

[glimchb]

Fixes #104

Signed-off-by: Boris Glimcher [email protected]

[codecov]

Codecov Report

Merging #156 (ef000a4) into main (297ebb8) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main     #156   +/-   ##
=======================================
  Coverage   75.00%   75.00%           
=======================================
  Files           5        5           
  Lines        1012     1012           
=======================================
  Hits          759      759           
  Misses        218      218           
  Partials       35       35           

[venkatmahalingam]

Please take the below change, the build should pass.

diff --git a/docker-compose.yml b/docker-compose.yml
index 1dd8df2..678da15 100644
--- a/docker-compose.yml
+++ b/docker-compose.yml
@@ -260,7 +260,7 @@ services:
       /entrypoint.sh call --json_input --json_output localhost:50151 CreateSvi            "{\"svi_id\"            : \"green-vlan40\",  \"svi\" :             {\"spec\" : {\"vrf\": \"//network.opiproject.org/vrfs/green\",  \"logical_bridge\": \"//network.opiproject.org/bridges/vlan40\", \"mac_address\" : \"qrvMAABB\", \"gw_ip_prefix\": [{\"addr\": {\"af\": \"IP_AF_INET\", \"v4_addr\": 673720321}, \"len\": 24}] }} }" && \
       /entrypoint.sh call --json_input --json_output localhost:50151 CreateSvi            "{\"svi_id\"            : \"yellow-vlan50\", \"svi\" :             {\"spec\" : {\"vrf\": \"//network.opiproject.org/vrfs/yellow\", \"logical_bridge\": \"//network.opiproject.org/bridges/vlan50\", \"mac_address\" : \"qrvMAABR\", \"gw_ip_prefix\": [{\"addr\": {\"af\": \"IP_AF_INET\", \"v4_addr\": 842150401}, \"len\": 24}] }} }" && \
       /entrypoint.sh call --json_input --json_output localhost:50151 CreateBridgePort     "{\"bridge_port_id\"    : \"eth1\",          \"bridge_port\" :     {\"spec\" : {\"mac_address\" : \"qrvMAAAB\", \"ptype\": \"TRUNK\",  \"logical_bridges\": [\"//network.opiproject.org/bridges/vlan10\", \"//network.opiproject.org/bridges/vlan20\", \"//network.opiproject.org/bridges/vlan40\"] }} }" && \
-      /entrypoint.sh call --json_input --json_output localhost:50151 CreateBridgePort     "{\"bridge_port_id\"    : \"eth2\",          \"bridge_port\" :     {\"spec\" : {\"mac_address\" : \"qrvMAAAB\", \"ptype\": \"TRUNK\",  \"logical_bridges\": [\"//network.opiproject.org/bridges/vlan50\"] }} }" && \
+      /entrypoint.sh call --json_input --json_output localhost:50151 CreateBridgePort     "{\"bridge_port_id\"    : \"eth2\",          \"bridge_port\" :     {\"spec\" : {\"mac_address\" : \"qrvMAAAB\", \"ptype\": \"ACCESS\",  \"logical_bridges\": [\"//network.opiproject.org/bridges/vlan50\"] }} }" && \
       echo get && \
       /entrypoint.sh call --json_input --json_output localhost:50151 GetVrf               "{\"name\" : \"//network.opiproject.org/vrfs/blue\" }" && \
       /entrypoint.sh call --json_input --json_output localhost:50151 GetVrf               "{\"name\" : \"//network.opiproject.org/vrfs/green\" }" && \
@@ -366,11 +366,10 @@ services:
                 ip route add default via 40.40.40.1 dev eth0.40 table 1001
                 sleep 5 && \
                 ping -c 3 6.6.6.6 -I 40.40.40.40 && ping -c 3 40.40.40.41 && \
-                ip link add link eth1 name eth1.50 type vlan id 50 && \
-                ip link set eth1.50 up && \
-                ip addr add 50.50.50.50/24 dev eth1.50 && \
+                ip addr add 50.50.50.50/24 dev eth1 && \
                 ip rule add from 50.50.50.50 lookup 1002
-                ip route add default via 50.50.50.1 dev eth1.50 table 1002
+                ip route add default via 50.50.50.1 dev eth1 table 1002
+                arp -s 50.50.50.1 aa:bb:cc:00:00:51
                 sleep 5 && \
                 ping -c 3 7.7.7.7 -I 50.50.50.50'

@@ -392,12 +391,6 @@ networks:
       config:
         - subnet: 10.168.3.0/24

-  internet:
-    ipam:
-      driver: default
-      config:
-        - subnet: 5.5.5.0/24
-
   internet1:
     ipam:
       driver: default
@@ -422,24 +415,12 @@ networks:
       config:
         - subnet: 10.10.10.32/28

-  n3htoleafbn1:
-    ipam:
-      driver: default
-      config:
-        - subnet: 20.20.20.0/24
-
   n4htoleafbn2:
     ipam:
       driver: default
       config:
         - subnet: 30.30.30.0/24

-  n5h1tol1r:
-    ipam:
-      driver: default
-      config:
-        - subnet: 40.40.40.16/28
-
   n6h1tol1y:
     ipam:
       driver: default

[JanScheurich]

The actual changes from TRUNK to ACCESS look OK.
Please consider my detailed comments.

[mardim91]

Do we need also to change the diagramm here @venkatmahalingam ?

https://github.com/opiproject/opi-evpn-bridge/blob/main/docs/OPI-EVPN-Leaf1-Detailed-View.png

[mardim91]

Is this arp entry on the arp table of host1 related to the mac addr of the GW ip ?

If yes why this is necessary ? Static arp entries should not be needed the host 1 should arp for the GW mac address automatically

[venkatmahalingam]

We dont need this anymore, please refer my above reply.

[mardim91]

I am curious why this is not happening with the Trunk case and only access case. I mean how the bridges on the host are reacting when they see packets with vlan tags on them taking into account that those bridges have no vlan configuration ? Do they drop the packets or they forward the packet based on the MAC dest ignoring the vlan tag ? Because I am curious how the virtual env works coreclty for trunk cases and only have problems for access case.
@venkatmahalingam ^

[glimchb]

removed static ARP ef000a4

[venkatmahalingam]

@mardim91 I'll check and get back for your questions.

[mardim91]

Ty @venkatmahalingam

[venkatmahalingam]

Do we need also to change the diagramm here @venkatmahalingam ?

https://github.com/opiproject/opi-evpn-bridge/blob/main/docs/OPI-EVPN-Leaf1-Detailed-View.png

Yes, @glimchb Please create issues for this and Jan's comment "It's a bit hard to correlate the bridge ports named "eth1" and "eth2" with the interfaces "eth0" and "eth1" of the opi-test container. Could you align those to make it easier to understand?" comment?

[mardim91]

lgtm

[glimchb]

Yes, @glimchb Please create issues for this and Jan's comment

opened #179 and #180