---------- Forwarded message ---------- From: LukeFlynn <lukehflynn(a)gmail.com&gt; Date: Mon, Sep 18, 2017 at 5:15 PM Subject: Re: [ovirt-users] PFSense VLAN Trunking To: FERNANDO FREDIANI <fernando.frediani(a)upx.com&gt; Good news, I found a way to trunk vlans! I couldn't get it working with native linux bridges, instead I opted to migrate my ovirt cluster to use OpenVSwitch instead (which is what XenServer is using among others). The default behavior of untagged OVS networks is to trunk all, so I simply had to create an untagged network on the same interface (or bond) as the tagged network and pass the untagged to pfsense. This didn't work with native bridging, but works as expected with OVS. Good news is that network throughput is great, I'm able to get full gigabit speeds! Just make sure you disable tx offloading in the PFSense web gui. There are a couple of gotchas when switching a cluster to OVS: 1.) It doesn't appear to be fully automated just yet (support is experimental) 2.) You must manually enable and start the openvswitch.service 3.) I had to manually tear down all of the network configuration VDSM did, and basically set the host networking up from scratch in the web UI after moving the cluster to OVS. Hope this helps some folks! Luke On Mon, Sep 18, 2017 at 8:26 AM, FERNANDO FREDIANI < fernando.frediani(a)upx.com&gt; wrote: