Score:0

Isolate traffic between vswitches but allow l2 traffic from external network

kz flag

I've two OVS bridges which they are connected to the 3rd v-switch using patch(es).

    Bridge br1
        Port br1
            Interface br1
                type: internal
        Port br1-ext
            Interface br1-ext
                type: patch
                options: {peer=ext-br1}
        Port veth573b5f8a
            Interface veth573b5f8a
    Bridge br2
        Port br2
            Interface br2
                type: internal
        Port vethe3fada17
            Interface vethe3fada17
        Port br2-ext
            Interface br2-ext
                type: patch
                options: {peer=ext-br2}
    Bridge ext
        Port physnet1
            Interface physnet1
        Port ext-br1
            Interface ext-br1
                type: patch
                options: {peer=br1-ext}
        Port ext
            Interface ext
                type: internal
        Port ext-br2
            Interface ext-br2
                type: patch
                options: {peer=br2-ext}

There's a DHCP server over physnet1. I need to isolate L2 traffic between these v-switches but allow DHCP traffic from physnet1. How i can achieve that?

Note: There's N v-switches IRL.

Score:0
kz flag

I found that I need to set rules in the flow table, I've achieved tenant isolation using these flows:

delete default flow on ext bridge and:


cookie=0x0, duration=84697.471s, table=0, n_packets=3486, n_bytes=987906, priority=100,in_port="ext-br1" actions=output:physnet1
cookie=0x0, duration=84687.469s, table=0, n_packets=1416, n_bytes=458415, priority=100,in_port="ext-br2" actions=output:physnet1
cookie=0x0, duration=84117.197s, table=0, n_packets=4884, n_bytes=1793355, priority=100,in_port=physnet1 actions=NORMAL
I sit in a Tesla and translated this thread with Ai:

mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.