net.md

Single node networking

This reference explains how network access to guests is achieved when deploying microstack using local access to the guests (normal for a single node deployment).

If the defaults are chosen sunbeam will have setup an external subnet within OpenStack which will be used to allocate floating IPs.

$ openstack subnet show -c allocation_pools -c cidr external-subnet
+------------------+-------------------------+
| Field            | Value                   |
+------------------+-------------------------+
| allocation_pools | 10.20.20.2-10.20.20.254 |
| cidr             | 10.20.20.0/24           |
+------------------+-------------------------+

The sunbeam installation adds a route to the machine that will route traffic for this range to the bridge br-ex.

$ sudo ip route
...
10.20.20.0/24 dev br-ex proto kernel scope link src 10.20.20.1 
...

In turn br-ex is configured with the first ip from the floating ip cidr (in this case 10.20.20.1), this IP is not included in the allocation range of the external-subnet.

$ sudo ip addr show br-ex
48: br-ex: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UNKNOWN group default qlen 1000
    link/ether 12:73:07:2f:9c:47 brd ff:ff:ff:ff:ff:ff
    inet 10.20.20.1/24 scope global br-ex
       valid_lft forever preferred_lft forever
    inet6 fe80::1073:7ff:fe2f:9c47/64 scope link 
       valid_lft forever preferred_lft forever

br-ex is then wired in the normal way with external access through br-ex and tap devices for the guests attached to br-int:

$ sudo openstack-hypervisor.ovs-vsctl show 
fbea0320-e3dd-49ca-8406-a15da3bb7347
    Bridge br-ex
        datapath_type: system
        Port patch-provnet-11a65cde-a791-4644-899d-fb9b28ab84ee-to-br-int
            Interface patch-provnet-11a65cde-a791-4644-899d-fb9b28ab84ee-to-br-int
                type: patch
                options: {peer=patch-br-int-to-provnet-11a65cde-a791-4644-899d-fb9b28ab84ee}
        Port br-ex
            Interface br-ex
                type: internal
    Bridge br-int
        fail_mode: secure
        datapath_type: system
        Port br-int
            Interface br-int
                type: internal
        Port tap44e1126e-d4
            Interface tap44e1126e-d4
        Port tap0c44773d-70
            Interface tap0c44773d-70
        Port patch-br-int-to-provnet-11a65cde-a791-4644-899d-fb9b28ab84ee
            Interface patch-br-int-to-provnet-11a65cde-a791-4644-899d-fb9b28ab84ee
                type: patch
                options: {peer=patch-provnet-11a65cde-a791-4644-899d-fb9b28ab84ee-to-br-int}
    ovs_version: "3.1.0"

There is also an iptables rule for the floating ip network to allow egress traffic from the quest.

$ sudo iptables-legacy -t nat -L POSTROUTING -n -v | grep -Ev 'cali|kubernetes'
Chain POSTROUTING (policy ACCEPT 53445 packets, 3302K bytes)
 pkts bytes target     prot opt in     out     source               destination         
   26  1896 MASQUERADE  all  --  *      *       10.20.20.0/24        0.0.0.0/0            /* openstack-hypervisor external network rule */