@LTS_Tom I have use the video bellow to setup vLAN and jails successfully in the past.
In my new homelab setup, I have create a LAGG interface with the 2 NIC that I have on the server. Could you please advice on how to create the vlan for the jails as I now only have 1 interface with an IP address on it.
Not something that I have done before, might want to check their fourms.
Will do. Thank you for the quick reply
I figured out somehow to make this work, not sure I could do it again, but I think I ended up having to put everything on VLANs and have no untagged traffic.
Place your physical ports in the LAGG, then add VLANS and subsequent Bridges as interfaces. The VLANs would have the LAGG as the parent interface with no bridge members, You can add IP addresses for access to the server. The corresponding bridges only have the specific VLAN as a bridge member. Your jails / VMs would have the bridge interfaces to have access to the appropriate VLANs. Be sure to have some type of other access to the server if you break the network connection.
Went through the excellent video a few times as well as the related post in TrueNAS forum and may look at doing LAGG too but before I run I’m looking to walk and have a couple of questions regarding the 2-NIC approach proposed in the video.
I have a 2.5GB NIC and two 1GbE NICs. I only connect/use one of the 1GbE NICs (dedicated to management traffic) so was thinking I could maybe put the second one to use for setting up VLANs coming/going from 2.5GbE NIC using the approach in the video
What I can’t find and don’t understand is if/how both NICs share (or not) the traffic coming in/out. If shared, is it in round robin-type fashion or other pattern? Or is it (ideally, actually) that it would (or could) only use the 1GbE as a fall back only if 2.5GbE was down? Also, does the 1GbE NIC used for this method of VLAN’ing need to be connected at all to my network for VLANs to work on 2.5GbE NIC?
As well, one thing I noticed when I added a second 2.5GbE NIC that didn’t turn out to work properly (but was recognized), was that it took over ‘re0’ and bumped my spare 1 GbE NIC, which was ‘re0’ before the new NIC, to ‘re1’. This suggests to me things could cease to work - or have to be reconfigured - if a NIC is added depending on how interface names are assigned. To avoid this, is there a way to “reserve” interface names, akin to DHCP reservation, using MAC addresses or something else?
Any info would be appreciated.
Not sure if you can do bonding with mixed NIC speeds, not something I have tested. For the methodology used in switch port link aggregation you choose the algorithm: 9. Network — FreeNAS®11.3-U2 User Guide Table of Contents
Thanks for the reply. So ignoring link aggregation, with two NICs of the same speed in a bridge, is the traffic simply split evenly or what? And how can I ensure my config doesn’t fail when I add a new NIC or one stops working? If I’m just not googling the right words, let me know what words I need to search for. Thanks.