On Mon, Jun 25, 2018 at 4:35 PM, Michael Watters <[email protected]> wrote:
> You should be able to use bonded interfaces with an IP on each VLAN > Interface for the ovirt hosts and the engine. For example, here is the IP > configuration for one of our VLANs. > > 10: bond3: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc > noqueue state UP group default qlen 1000 > link/ether 00:1b:21:5c:80:39 brd ff:ff:ff:ff:ff:ff > inet6 fe80::21b:21ff:fe5c:8039/64 scope link > valid_lft forever preferred_lft forever > 24: bond3.311@bond3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc > noqueue state UP group default qlen 1000 > link/ether 00:1b:21:5c:80:39 brd ff:ff:ff:ff:ff:ff > inet 192.168.111.201/24 brd 192.168.111.255 scope global bond3.311 > valid_lft forever preferred_lft forever > inet6 fe80::21b:21ff:fe5c:8039/64 scope link > valid_lft forever preferred_lft forever > > bond3.311 configuration is managed in the > /etc/sysconfig/network-scripts/ifcfg-bond3.311 > file. > IMO setting up bonded NICs with VLAN tagging is one area where ovirt falls > short. You essentially have to configure your networks twice. First using > the /etc/sysconfig/network-interface files and then inside of the engine > itself. > I am not familiar with this problem, a vlan network over a bond is supported. I even recall several fixes related to it in 4.2. > VDSM may also need to be configured to use ifcfg persistence in the config > file. > > cat files/vdsm/vdsm.conf > [vars] > ssl = true > net_persistence = ifcfg > > The ifcfg persistence mode is planned for deprecation in 4.3, and is not well tested/supported for some time now. Do not use it unless you have a very (very) good reason to do so. > > [addresses] > management_port = 54321 > > > Your switch ports also need to be configured to support 802.1q networking. > > > > On 06/23/2018 09:56 AM, [email protected] wrote: > > Hi, > > > > I setup oVirt a few years back… now that the HCI is real, I am revisiting > it. I have deployed with Gluster, and am now moving on to networking. > > > > I come from a VMware shop and normally we trunk all the network ports > exposing all VLANs to the hosts and place VMs in Portgoups that are tagged > with VLANs. > > > > I did manage to do this years back but I am struggling to get this to work > today. I had pretty limited hardware back then and I thought I installed > using vlan tagging and trunked ports, but I don’t see any option to do > this using the glusterfs and hosted engine setup. > > > > Each host has a dual port 10Gb NICs I use one for storage that is > connected to my storage network and one for ovirtmgmt. (I need to add > another of these for redundancy down the road, but no money for that at the > moment) > > > > The hosts also have 4 x 1Gb ports. So in lieu of being able to configure > vlan tagging to trunked ports on hosted engine deploy, I am considering > cabling up a 1 Gb port on each in my management services VLAN and when it > is all up and running create another logical network (or several of them as > I think these equate to what is a vlan tagged port group in VMware) with > the 10Gb NIC backing for VMs. > > > > Does that sound reasonable? Or if anyone can point me to any docs that > describe how to deploy to a specific VLAN with trunk ports, that would be > nice as well as I won’t have to actually go to the office and run > additional cables. > > As mentioned by Michael, you can use vlan networks and attach several of them on the same bond/nic in the setup networks window. You also have the option to use a non-vlan network, and define the vlans inside the VM. The linux bridge that connects the vnics to the bond ignore tagging information, it just forwards frames in a flat manner, leaving it to the vnic in the VM to strip the tag. > > > Appreciate any advice > > > > Bill > > > > > _______________________________________________ > Users mailing list -- [email protected] > To unsubscribe send an email to [email protected] > Privacy Statement: https://www.ovirt.org/site/privacy-policy/ > oVirt Code of Conduct: > https://www.ovirt.org/community/about/community-guidelines/ > List Archives: > https://lists.ovirt.org/archives/list/[email protected]/message/TDFK2BNP4DMBB22SBNEXQLTOK5SOW42O/ > > > > _______________________________________________ > Users mailing list -- [email protected] > To unsubscribe send an email to [email protected] > Privacy Statement: https://www.ovirt.org/site/privacy-policy/ > oVirt Code of Conduct: https://www.ovirt.org/community/about/community- > guidelines/ > List Archives: https://lists.ovirt.org/archives/list/[email protected]/ > message/4NVTFK2I5LOWCS4XLSCP5BEERS7NXXNR/ > >
_______________________________________________ Users mailing list -- [email protected] To unsubscribe send an email to [email protected] Privacy Statement: https://www.ovirt.org/site/privacy-policy/ oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/[email protected]/message/SYPYJAMVHHQJHT2X6LTOCH3QOUNTO65Z/
