8:56 a.m.
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.
Appreciate any advice
Bill
8:35 a.m.
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. 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
[addresses]
management_port = 54321
Your switch ports also need to be configured to support 802.1q networking.
On 06/23/2018 09:56 AM, william.dossett(a)gmail.com 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.
Appreciate any advice
Bill
_______________________________________________
Users mailing list -- users(a)ovirt.org
To unsubscribe send an email to users-leave(a)ovirt.org
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/users@ovirt.org/message/TDFK2BNP4DM...
2:11 a.m.
On Mon, Jun 25, 2018 at 4:35 PM, Michael Watters <wattersm(a)watters.ws>
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, william.dossett(a)gmail.com 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 -- users(a)ovirt.org
To unsubscribe send an email to users-leave(a)ovirt.org
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/users@ovirt.org/message/TDFK2BNP4DM...
_______________________________________________
Users mailing list -- users(a)ovirt.org
To unsubscribe send an email to users-leave(a)ovirt.org
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/users@ovirt.org/
message/4NVTFK2I5LOWCS4XLSCP5BEERS7NXXNR/
9:31 a.m.
just got back to this... the method I usee worked well.
I used a 1GB copper port in my mgmt VLAN to set up oVirt with Gluster
Got it all up and running then used 1 x 10 GB port to create my storage network
I think created to logical networks, one tagged with my static VLAN and the other tagged
with my dhcp VLAN, dragged them both onto the second 10Gb port which is connected to cisco
FEX trunk port - worked perfectly and I can deploy to either VLAN now. I guess I will
keep the 1Gb for management... I think I can transfer it to the 10Gb VLANs most likely,
but didn't want to lose my ovirt engine as this setup seems to be working well for me
at least on the network side.
2107
days inactive
2141
days old
3 comments
3 participants
participants (3)
-
Edward Haas -
Michael Watters -
william.dossett＠gmail.com